HCA2137/2004
HCA 1383/2012
IN THE HIGH COURT OF THE
HONG KONG SPECIAL ADMINISTRATIVE REGION
COURT OF FIRST INSTANCE
ACTION NO 1383 OF 2012
--------------------
BETWEEN
DR YEUNG, SAU SHING ALBERT Plaintiff
and
GOOGLE INC Defendant
---------------------
Before: Deputy High Court Judge Marlene Ng in Chambers
Date of Hearing: 8 May and 7 November 2013
Date of Handing Down Decision: 5 August 2014
------------------------
D E C I S I O N
------------------------
Index Page
I. Introduction 2
II. Summons 7
III. Issues 9
IV. Autocomplete and Relates Search Features 10
V. Service out of jurisdiction 15
VI. Good arguable case: publication 15
VII. Good arguable case: publisher 22
a) Law 24
b) Facts 59
c) Striking a balance 69
d) Summary 77
VIII. Serious question to be tried 77
a) Real and substantial tort 77
or Jameel abuse of process
b) .com and .com.tw Websites 85
c) Scale of publication 89
d) Alternative claims 96
e) Previous litigation 99
f) Summary 99
IX. Conclusion 100
I. INTRODUCTION
1. Technological developments and innovations have rendered our lives more comfortable and easier. The worldwide web (internet) is a remarkable example of this development, and “googling” has become a popular mode of informing gathering. But such usage can result in both comfort and danger. Left unmonitored, there may be disquieting consequences through misinformation and unaddressed complaints, especially when the search process throws up unsavory information associated with a person or entity that proves to be derogatory and false. Questions have arisen in international legal forums over this troubling grey area (the combined diligence of counsel have presented me with over 50 authorities), and now at issue in this case in Hong Kong are the twin questions of who is responsible for predictive suggestions generated by the Autocomplete and Related Searches search engine features, and whether defamation law principles are amenable to address such concern.
2. The plaintiff (“Yeung”) is a businessman and chairman/founder of the Emperor Group being a group of listed companies based in Hong Kong and engaged in various business enterprises (“Group”), including property development and investment, retailing, hospitality and gaming, financial services, entertainment and movies as well as publishing and distribution. Yeung is involved in community services, and responsible for establishing various charitable funds and foundations. He has received recognition for his charity work.
3. The defendant (“Google Inc”) is a global business with headquarters in the United States that provides internet based services. Such services include internet search facilities through the use of their internet search engine known as “Google Search” (“Google Search”). Google Search is accessible by entering the name of any of the worldwide domains owned by Google Inc into a web browser. Google Inc is the owner, provider, administrator and manager of the domains/websites at .hk, .tw and (“.com.hk Website”, “.com.tw Website” and “.com Website”, collectively “Websites”). The .com Website is a top level domain that comprises the top level domain Google Inc uses for their services in the United States, and the .com.hk and .com.tw Websites that are the country code top level domains for services provided by them in Hong Kong and Taiwan respectively.
4. Yeung claims that on 23 May 2012, a search was conducted on the Websites as follows:
a) When the following Chinese characters and/or English words were typed in the “Google Search”/“Google 搜尋” search box (“Search Box”) on the homepage of the Websites, Google Search instantaneously and automatically generated a list of Autocomplete search suggestions in the drop-down menu for the user’s selection even before the user clicked the search button, some of which suggestions contained the following characters and/or words:
(i) typing “楊受成” generated “楊受成 黑社會”, “楊受成 變態” and “楊受成 新義安”[1] (“1st Words”);
(ii) typing “albert yeung” and/or “albert yeung sau shing” generated “albert yeung triad” and “albert yeung sau-shing triad” (“2nd Words”).
b) When the following Chinese characters and/or English words were typed in the Search Box on the homepage of the Websites and the search button was clicked, Google Search generated at the bottom part of the search outcome/results page a list of “Related Searches”/“相關搜尋”, some of which contained the following characters and/or words:
(i) typing “楊受成” generated “楊受成 黑社會”, “楊受成十宗罪”, “楊受成 新義安”, “楊受成 14k”, “楊受成 內幕” and “容祖兒楊受成內幕”[2] (“3rd Words”);
(ii) typing “albert yeung” generated “albert yeung sau-shing triad” (“4th Words”).
5. Yeung claims Google Inc published or caused to be published the 1st, 2nd, 3rd and 4th Words (collectively, “Words”) that are defamatory of and concerning him during the period from 23 to 28 May 2012. For convenience, in this Decision I shall refer to the three search queries/terms of “楊受成”, “albert yeung” and/or “albert yeung sau shing” as the “Keywords”.
6. By a letter dated 28 May 2012 (“1st Letter”), Yeung’s solicitors (“ILL”) (a) notified Google Inc of the presence of the Words published on the Websites, (b) claimed that although Google Inc’s Autocomplete and Related Searches algorithms “may not involve any human intervention, they nevertheless generate and display information/ suggestion to internet users” which in the present case contained words/phrases that were defamatory of and concerning Yeung, and (c) demanded Google Inc to remove/prevent the Words from appearing/reappearing in any existing/future “google” search made against the Keywords. Yeung claims that by reason of the 1st Letter Google Inc knew or ought to have known of the presence of the Words on the Websites at least from 28 May 2012.
7. On 21 June 2012, Yeung through ILL wrote to Deacons (former solicitors for Google Inc in previous litigation between Yeung and Google Inc)[3] to draw their attention to the presence of offending words in the autocompletions upon typing the Keywords in the Search Box of the Websites, and to enquire whether they had instructions to act in the matter (“2nd Letter”). By the 2nd Letter, Yeung also indicated his wish to explore the possibility of an amicable resolution of the matter similar to the eventual disposal of the earlier litigation (ie removal of the defamatory material by Google Inc and discontinuance of the legal action).
8. Yeung claims that notwithstanding notification by the 1st and 2nd Letters Google Inc has failed to take down the Words and has continued to publish or cause to be published the Words at the Websites from 28 May 2012 (the date of the 1st Letter) to 6 August 2012 (the date of the Writ of Summons referred to in the paragraph below) (“Period”).
9. On 6 August 2012, Yeung issued a Writ of Summons against Google Inc (“WoS”) for inter alia damages for libel in respect of the publication of the Words that are defamatory of and concerning him in the manner set out in paragraph 4 above during the Period and continuing, and an injunction order restraining Google Inc from howsoever publishing and/or causing the publication of and/or participating in the publication of the alleged libels.
10. Yeung claims his personal and/or business reputation has been gravely injured by reason of the publication of the Words, and he relies on the following facts and matters in aggravation of damages:
a) the publication of the Words relates to allegations against him that are of a criminal nature and/or otherwise serious;
b) Google Inc has failed to take any or any reasonable step to verify the truth/accuracy of the Words, which contain allegations against him that are of a criminal nature and/or otherwise serious;
c) despite receipt of the 1st Letter, Google Inc has continued to publish or caused to be published the Words on the Websites, thereby causing further harm and suffering to him.
11. On 16 August 2012, Yeung filed the affirmation of Hom Mun Yee Caroline (“Hom”)[4] in support of his ex parte application for leave to serve the WoS on Google Inc out of jurisdiction in the United States pursuant to Order 11 rule 1(1)(f) of the Rules of the High Court (“RHC”) (“Hom 1st Aff”).
12. On 28 August 2012, ILL wrote to the court to request for the issue of a concurrent WoS. On 3 September 2012, Master K Lo inter alia granted leave for Yeung to issue/serve a concurrent WoS against Google Inc out of jurisdiction in the United States (“Leave Order”). On 14 September 2012, the concurrent Writ of Summons (“CWoS”) was issued out of court.
13. On 4 October 2012, Google Inc’s solicitors (“Deacons”) filed acknowledgment of service giving notice of intention to defend. On 5 November 2012, Yeung filed his Statement of Claim (“SoC”).
II. SUMMONS
14. On 24 December 2012, Google Inc issued a summons under Order 12 rule 8 of the RHC and/or the inherent jurisdiction of the court for inter alia the following reliefs (“Summons”):
a) a declaration that the court has no jurisdiction over Google Inc in respect of the subject matter of the claim or the reliefs/ remedies sought against Google Inc in the present action, or alternatively a declaration that the court should not exercise any jurisdiction it may have;
b) an order that (i) the CWoS and Leave Order be set aside or discharged, (ii) service of the CWoS on Google Inc be set aside, and (iii) all subsequent proceedings in the present action be stayed.
15. As set out in the Summons, the grounds that Google Inc relies are (a) the High Court of Hong Kong does not have personal jurisdiction over them, (b) there is no good arguable case that Yeung’s claim falls within Order 11 rule 1(1)(f) of the RHC, and (c) there is no serious issue to be tried on the merits of Yeung’s case. Google Inc also relies on matters further set out in the affidavit of their solicitor Robert George Clark (“Clark”) filed on 24 December 2012 (“Clark 1st Aff”). There is an implicit allusion to a complaint of material non-disclosure in the Clark 1st Aff, but Clark makes an express complaint in his affidavit in reply filed on 17 April 2013 (“Clark 2nd Aff”).
16. On 20 February 2013, Yeung filed Hom’s 2nd affirmation and the affirmation of Yuen Yin Man Timothy (“Yuen”)[5] in opposition (“Hom 2nd Aff” and “Yuen Aff”).
17. For the purpose of the Summons, Google Inc does not take issue over whether the Words are defamatory of Yeung.
III. ISSUES
18. As explained above, Yeung’s application for the Leave Order is based on Order 11 rule 1(1)(f) of the RHC. The Hom 1st Aff contends that Yeung’s claim is founded on the tort of defamation, and either the damage has been sustained within this jurisdiction or the damage has resulted from an act committed within the jurisdiction.
19. The Hom 1st Aff states that as a matter of law the material part of the cause of action in libel is the publication of the libel, and in order to constitute publication the matter must be published to a third party. It goes on to say the tort is committed at the place where the publication is received by the reader or viewer, and in relation to electronic publication, it takes place where a person downloads the material and not where the material was uploaded.[6]
20. Mr McCoy SC, counsel for Google Inc, accepts the above principles are prima facie uncontroversial, but suggests that Yeung has only partly fulfilled his duty by merely drawing attention to these principles at the ex parte stage. He submits there is no evidence of publication of the Words to a genuine third party in Hong Kong and/or Yeung has failed to highlight to the master the principles/authorities that show Google Inc is not a publisher of the predictions (including the Words) or even the search results themselves which Google Search generates without human intervention. Hence, there is no good arguable case of Yeung’s claim falling within Order 11 rule 1(1)(f) of the RHC, and no serious question to be tried on the merits to enable the court to exercise its discretion to grant the Leave Order. Although not explicitly mentioned in the Summons,[7] Mr McCoy SC also suggests there has been material non-disclosure at the ex parte stage that justifies discharge of the Leave Order.
21. On the other hand, Mr Ng, counsel for Yeung, says that Google Inc as a search engine is liable for failing to take down the defamatory Words they as publisher have published on the Websites in Hong Kong to third parties after having been notified of the presence of such Words on the Websites. Mr Ng disagrees there has been any material non-disclosure, and submits there is a good arguable case that Yeung’s claim is within Order 11 rule 1(1)(f) of the RHC as well as a serious question to be tried.
IV. AUTOCOMPLETE AND RELATED SEARCHES FUNCTIONS
22. Google Inc owns and operates Google Search. The internet has billions of publicly indexable webpages and billions of internet users. Every day Google Search answers more than one billion queries in 181 countries and 146 languages. To understand Yeung’s complaints in relation to Google Search’s Autocomplete and Related Searches features, it is necessary to detail how they work.
23. If an internet user knows a specific website address, he/she can visit it without using a search engine. But to find information that is not presented to the user as a specified website address or an active link, a search engine is his/her portal to the internet. It enables the user to peruse a huge volume of websites which would otherwise be an unmanageable task.
24. Google Search has automated information retrieval and index construction systems that crawl the web and index the webpages. Its web crawling bot or “spider” is called Googlebot. As Googlebot crawls the web, it processes the webpages and compiles a massive index (similar to that of a large book) of all the words it sees and their locations on each webpage. New sites, changes to existing sites and dead links are noted, and the new and updated data is added to the index. Googlebot uses complicated and sophisticated algorithmic processes to decide which websites to crawl, how often, and how many pages to fetch from each website. Websites that want to block Googlebot from crawling their content can take steps to do so, including using robots.txt to block access to files and directories on their servers.
25. When a user performs a typical “google” search, the following information flow occurs: (a) the user inputs a search query by typing the desired search term(s)/keyword(s) into the web browser, (b) Google Search displays or serves the results to the query, and (c) the user receives the content.
26. Traditionally, the user types his search query in the Search Box and clicks the search button, then Google Search begins the search process by crawling and fetching webpages. Googlebot’s automated and pre-programmed algorithmic search processes extract information from its index and find matching webpages to return results that contain or are relevant to the term(s)/keyword(s) entered by the user. This is akin to a librarian looking through the index cards of the library trying to identify relevant books for a library user who wishes to read books on a particular topic, but a “google” search does so electronically on a gargantuan scale. Typically, the search results generated by Googlebot contain the name of the identified piece of content, a hyperlink to its location, and a snippet/excerpt from the webpage that shows how the content relates to the search query.
27. Google Search’s ranking algorithms use over 200 unique signals to determine which of the webpages out of the many are most relevant to the search query according to popularity, demand and/or usage. One such signal is PageRank which measures the importance of a webpage based on the number/quality of incoming links from other pages. Other factors include how often the search terms occur on the page, if they appear in the title or whether synonyms of the search term(s) occur on the page. Google Inc touts the outcome of its crawl and search relevancy algorithms as presenting “the most relevant search results towards the top of the page, and the less relevant search lower down the page”.
28. Google Search’s Autocomplete functionality takes a step beyond the traditional approach by allowing Google Search to operate with greater relevance and efficiency to its users by helping them to find information quickly. As the user types his search query in the Search Box, Google Search begins to offer predictive suggestions or keyword generations in real-time in the drop-down menu to finish the keyword phrase. The Autocomplete feature displays constantly updated search suggestions for the user to choose, and the predictive suggestions change on each additional keystroke as the user types in the Search Box.
29. Google Search’s Related Searches feature is a search engine tool that helps the user to quickly look for results similar to his search query that he may be missing or if he is not sure exactly what he is looking for. When a user types search term(s)/keyword(s) in the Search Box, the Related Searches feature displays popular refinements that are commonly associated with the original search query and/or are slightly more detailed so the user can narrow down his search. The search options generated by the Related Searches feature are displayed at the bottom of the search outcome/results page, and the user can click/select any of the variations so displayed in order to dig deeper into that topic.
30. Underlying the Autocomplete and Related Searches features, which are designed by Google Inc to increase search efficacy, is Google Search’s complex suggested search and related terms algorithms that reflect past search activity of all “google” search web users (which in turn reflect the popularity and volume of the search queries they have previously typed) and the content of the webpages indexed by Google Search. It is the previous search activity and web content that drive the suggestions.
31. The relevancy and popularity of Autocomplete predictions and Related Searches results produced by Google Search’s algorithms and computer programmes may alter upon changes in user search activity and internet content. Search terms as applied by internet users may change over time, and internet content changes with new webpages being introduced and old webpages being edited or taken down. A user typing “McDonalds” in the Search Box may throw up “McDonalds hamburger” as an autocompletion on a given day and “McDonalds cheeseburger” on another occasion. However, to influence or change what the autocomplete instant results show will require a large number of users with unique internet protocol (IP) addresses to type the desired search query into Google Search on an ongoing basis.
32. In face of the operations described above, Google Inc contends that it is not a “publisher” of the results of Google Search’s Autocomplete and Related Searches functionalities. Mr McCoy SC argues that by providing internet search services through Google Search’s automated search algorithms without human input, Google Inc is a mere passive facilitator in respect of the words/images seen on its domains.
33. For the purpose of the Summons, Yeung does not dispute the automated nature of the Google Search processes, but Mr Ng argues that through their design, maintenance and improvement of the algorithms used to (a) compare search queries with Googlebot’s index, (b) decide the relative importance of webpage ranking, and (c) reconstitute web content and aggregate previous search activity, Google Inc “intended” and in this sense has control over the Autocomplete and Relates Searches results that appear on the Websites, and is therefore a “publisher” of such results.
34. I will return to these matters below.
V. SERVICE OUT OF JURISDICTION
35. In general, to obtain leave to serve a writ of summons out of jurisdiction, a plaintiff is required to:
a) show in relation to his claim he has a good arguable case that it falls within the ground relied upon under Order 11 rule 1(1) of the RHC;
b) satisfy the court there is a serious question to be tried on the merits of the claim, ie a substantial question of fact or law or both;
c) satisfy the court that in all the circumstances, Hong Kong is clearly and distinctly the appropriate forum for the trial of the dispute, and the court ought to exercise its discretion to permit service of the writ of summons out of the jurisdiction.
VI. GOOD ARGUABLE CASE: PUBLICATION
36. Each person who publishes defamatory material is potentially liable. Question therefore arises as to whether Google Inc can be regarded as a “publisher” of the Words being predictions or suggestions derived from Google Search’s search processes, and if so, whether Google Inc has published the Words to a third party publishee. I shall deal with the latter question first.
37. Publication is a bilateral act. In defamation cases, material is “published” and damage occurs where the publication is comprehended by the reader, and it ordinarily occurs at the place where the material is made available. In the internet context, as explained above, it is at the place where the material is viewed/downloaded (rather than where it was posted/ uploaded onto the internet) provided that the claimant has a reputation capable of being damaged in the jurisdiction in which he wishes to bring his action.[8] Thus, an internet publisher who places material on the internet will be responsible for the effects of his action whenever the damage occurs.
38. Here, the Hom 1st Aff confirms that when the 1st Letter was published, the Group’s information and technology department (“IT Dept”) was able to download and print the Words from the Websites, and further contends that such users are plainly different from Yeung. Further, when there was no response to the 1st Letter and Google Inc still did not remove the Words from the Websites, ILL downloaded the Words from the Websites on 6 August 2012 before issuing the WoS on the same day.
39. Mr McCoy SC has no quarrel with the principles set out in paragraph 37 above, but says that without actual evidence of a genuine third party user conducting a search using the Keywords and giving evidence of the precise results thrown up by the Autocomplete and Related Searches features that such user has seen on the web browser at a particular time/date, there can be no certainty of what has been seen or “published”. Since the users from the IT Dept and/or ILL are “persons employed by or tasked by [Yeung] to deliberately search, find and ‘publish’ material allegedly defamatory of him”, there is no third party user and hence no publication.
40. Mr McCoy SC submits that for well over 150 years since The Duke of Brunswick v Harmer,[9] the common law has turned its face away from an artificial and orchestrated situation whereby a claimant sets his servant or agent (in contra-distinction to a genuine third party) to read the libel as constituting publication to a third party. He says that at the ex parte stage Yeung has failed to provide evidence of publication of the Words to a true third party in Hong Kong, and in failing to raise this issue in the Hom 1st Aff he is also in breach of his duty to make material disclosure.
41. I disagree. In The Duke of Brunswick, the Duke was given a copy of a newspaper published 17 years previously that contained material defamatory of him. Such newspaper was sold to a person sent by the Duke to procure it. But contrary to Mr McCoy SC’s submissions, Coleridge J held that:
“…… the defendant, who, on the application of a stranger, delivers to him the writing which libels a third person, publishes the libellous matter to him, though he may have been sent for the purpose of procuring the work by that third person. So far as in him lies, he lowers the reputation of the principal in the mind of the agent, which, although that of an agent, is as capable of being affected by the assertions as if he were a stranger. The act is complete by the delivery: and its legal character is not altered, either by the plaintiff’s procurement or by the subsequent handing over of the writing to him. Of course that this publication was by the procurement of the plaintiff is not material to the question we are now considering.”
42. This is also borne out by the judgment by Lord Esher MR in Pullman v Walter Hill & Co Ltd[10] cited with approval in Godfrey v Demon Internet Ltd:[11]
“The first question is, whether assuming the letter to contain defamatory matter, there has been a publication of it. What is the meaning of ‘publication’? The making known the defamatory matter after it has been written to some person other than the person of whom it is written … If the writer of a letter shows it to his own clerk in order that the clerk may copy it for him, is that a publication of the letter? Certainly it is showing it to a third person; the writer cannot say to the person to whom the letter is addressed, ‘I have shown it to you and to no one else.’ I cannot, therefore, feel any doubt that, if the writer of a letter shows it to any person other than the person to whom it is written, he publishes it. If he wishes not to publish it, he must, so far as he possibly can, keep it to himself, or he must send it himself straight to the person to whom it is written. There was therefore, in this case a publication to the type-writer.”[12]
43. Mr McCoy SC submits that Jameel (Yousef) v Dow Jones & Co Inc[13] held that where the plaintiff procures publication to his agent or servant for the purpose of defamation proceedings, as in the present case, it amounts to an abuse of process, and he relies in particular on the following observations:
“55. There have been two recent developments which have rendered the court more ready to entertain a submission that pursuit of a libel action is an abuse of process. The first is the introduction of the new Civil Procedure Rules. Pursuit of the overriding objective requires an approach by the court to litigation that is both more flexible and more proactive. The second is the coming into effect of the Human Rights Act 1998. Section 6 requires the court, as a public authority, to administer the law in a manner which is compatible with Convention rights, insofar as it is possible to do so. Keeping a proper balance between the article 10 right of freedom of expression and the protection of individual reputation must, so it seems to us, require the court to bring to a stop as an abuse of process defamation proceedings that are not serving the legitimate purpose of protecting the claimant's reputation, which includes compensating the claimant only if that reputation has been unlawfully damaged.
56. We do not believe that Brunswick v Harmer 14 QB 185 could today have survived an application to strike out for abuse of process. The Duke himself procured the republication to his agent of an article published many years before for the sole purpose of bringing legal proceedings that would not be met by a plea of limitation. If his agent read the article he is unlikely to have thought the Duke much, if any, the worse for it and, to the extent that he did, the Duke brought this on his own head. He acquired a technical cause of action but we would today condemn the entire exercise as an abuse of process.” (my emphasis)
44. I am unable to discern from the above passage support for Mr McCoy SC’s proposition. In Jameel (Yousef), the court held that an action can be struck out on the ground that a libel claim with minimal publication that caused no real damage within the jurisdiction is, in the circumstances, an abuse of process. What this case demonstrates, at least in England and Wales, is that a defendant who fails to challenge or who has unsuccessfully challenged jurisdiction is not precluded from later advancing an argument of minimal publication and/or insubstantial damage to support an application to strike out the claim as an abuse of process. It is in this vein that the English Court of Appeal suggested The Duke of Brunswick would not have survived today, but they nevertheless accepted the Duke did “[acquire] a technical cause of action”. As explained in Gatley, “[since] publication to one person will suffice, it is not necessary that there should be a “publication” in the commercial sense, though the scale of the publication will of course affect the damages. …… However, where the publication is to a small number of people any claim risks being struck out as an abuse of process ……”[14] I will return below to the issue of Jameel abuse of process, but suffice to state here that this case does not support Mr McCoy SC’s propositions as set out in paragraphs 40 and 43 above.
45. Mr McCoy SC submits there is already analogous law in Hong Kong that supports his propositions albeit in a different context. He claims it was held in Oriental Press Group Ltd v Next Magazine Publishing Ltd & ors,[15] Oriental Press Group Ltd v Next Magazine Publishing Ltd & ors (No 2)[16] and Oriental Press Group Ltd v Next Magazine Publishing Ltd & ors[17] that the particular person chosen by the plaintiff as the reader of the impugned article and as the person with specific knowledge of additional facts that would give rise to an inneuendo meaning could not be regarded as a member of the public in the true sense as required under the law of defamation.
46. As rightly pointed out by Mr McCoy SC, the authorities cited in the above paragraph concern a wholly different context. In that case, the plaintiff pleaded its defamation claim in the alternative by relying on the natural and ordinary meaning of the impugned article within the general knowledge of ordinary readers or a defamatory meaning that arises because of extrinsic facts known only to the recipients. The latter, being a plea of “true” or “legal” innuendo, has two principal consequences: (a) the plaintiff must not only identify in its particulars of claim the defamatory meaning which it contends the words convey but also identify the relevant extrinsic facts and prove those facts were known to at least one of the persons to whom the words were published, and (b) the meaning resulting from those facts gives rise to a cause of action separate from that (if any) arising from the words in their ordinary and natural meaning because it is an extended meaning not present in the words themselves.[18] Hence, the extended innuendo meaning must arise from facts passing beyond general knowledge that are “within the knowledge of the persons to whom the words were published”.[19]
47. In that case, the learned judge held that the publisher of Oriental Daily News, who was an employee of the plaintiff for many years and whose knowledge/understanding of the plaintiff’s structure and the composition of their management level, could not be equated with that of an ordinary member of the public, still less within any class of ordinary readers of Next Magazine to whom the article was published. But I cannot see how the magazine’s readership relevant for ascertaining the meaning and innuendo of the alleged defamatory material has any bearing on the question of whether or not there is any act of publication that does not depend on any targeted readership but simply on any person other than the claimant who has viewed the impugned material.
48. In my view, there is plainly a good arguable case that viewing, downloading and printing the Words from the Websites by users at the IT Dept and ILL constitute publication of the Words to third party publishees.
VII. GOOD ARGUABLE CASE: PUBLISHER
49. Mr McCoy SC submits that at law Google Inc is not a publisher since no human input/operation is required in the search processes for the Autocomplete and Related Searches results to appear. As a search engine, Google Inc is a mere passive medium of communication without any mental element. He says Yeung’s claim is therefore wholly unarguable.
50. On the other hand, Mr Ng submits that even if the Google Search system is automated, Google Inc actively facilitates the publication of the Words. They intend Google Search to operate and to publish the results (including the Words) generated by its automated systems when a “google” search is made, which is what such systems are designed to do. Mr Ng says that when the Words on the Websites are viewed or downloaded by an internet user, there is, as intended by Google Inc, publication by them as owner and operator of Google Search. The Words are not trespassing inscriptions by a third party on Google Inc’s property (ie the Websites) without their approval or encouragement such that they can invoke the principle in Bryne v Deane.[20]
51. Mr Ng argues that even if the Bryne principle is applicable, the courts in common law jurisdictions have consistently held that liability will be established against an internet service provider (“ISP”) or search engine if, after having been notified of the presence of defamatory material on their server/website, they fail to remove it, and such liability is premised on their knowledge of and/or control over the defamatory material. In the present case, the 1st Letter (reinforced by the 2nd Letter) has brought home to Google Inc the presence of the Words on the Websites as well as Yeung’s complaint about them, and the “google” search by ILL using the Keywords on 6 August 2012 demonstrates the continued presence of the Words on the Websites even after notification.
52. Alternatively, Mr Ng contends that whether Google Inc is a publisher raises significant and complex legal issues, and it is inappropriate to decide such controversial and developing area of the law in an application to set aside service out of jurisdiction without eliciting all the facts and evidence.
53. Mr McCoy SC says there is no case law he can find in the common law jurisdictions on liability for defamation in respect of autocompletions and related searches results, and Yeung is not entitled to seek the Leave Order on a novel point by stretching existing case law on the general liability for libel in respect of search engines. He urges the court to bear in mind that Yeung’s claim as presently framed is a narrow one relating to autocompletions and related search suggestions, and on such refined basis the current development of case law does not support Yeung’s contentions.
54. I accept that the electronic technology revolution has brought new challenge to the orthodox legal framework for libel actions, but it is equally true that the success of the common law derives from the fact that it is responsive to change. As Sir James Munby P said in Re J (A Child) (albeit not in a libel context):[21]
“…… The law must develop and adapt, as it always has done down the years in response to other revolutionary technologies. We must not simply throw up our hands in despair and moan that the internet is uncontrollable. Nor can we simply abandon basic legal principles. ……”
55. With such thought in mind, I turn to the issue of whether or not there is a good arguable case that Google Inc is a publisher of the Words on the Websites. This is a matter of mixed fact and law,[22] and I shall start by turning first to the law.
56. I note at the outset both Mr McCoy SC and Mr Ng confirm that for the purpose of the Summons they will not rely on authorities from the civil law jurisdictions and from Japan cited in their written submissions.
(a) Law
57. As explained above, the gist of a defamation claim is not in the creation of the libel but in its publication. Ribeiro PJ in Oriental Press Group Ltd & anor v Fevaworks Solutions Ltd said as follows:[23]
“18. As Issacs J explained, “To publish a libel is to convey by some means to the mind of another the defamatory sense embodied in the vehicle …” ……
19. Until mitigated by the common law defence of innocent dissemination which evolved in the late nineteenth century, liability for publishing a libel was strict and could lead to harsh results. A person was held liable for publishing a libel if by an act of any description, he could be said to have intentionally assisted in the process of conveying the words bearing the defamatory meaning to a third party, regardless of whether he knew that the article in question contained those words. ……”
Consequently, any person who takes part in making the defamatory statement known to others may be liable for it. So under the strict publication rule, publication of a libel by, say, a newspaper means the journalist (who is originator of the article), the editor (who accepts/prepares it for publication), the printer (who sets the type and prints it), the wholesale distributor (who disseminates it), the newsagents (who sell it to the readers), and the newspaper’s proprietor (who published it through its employees or agents) are all liable.[24]
58. The harshness of the strict publication rule has led to some relaxation by the development of the common law defence of innocent dissemination, but such defence is available only to a subordinate (not main) publisher who does not know and will not with reasonable care in the relevant circumstances have known the article contains defamatory content, eg a newsagent who has sold a consignment of newspapers that contains an offending article may rely on such defence if he is not aware of the defamatory content and cannot, with reasonable care, have discovered it.[25]
59. Such defence that may be open to a subordinate publisher (ie a publisher who in the ordinary course of business plays subordinate role in the process of disseminating the impugned article, eg persons such as wholesalers, distributors, booksellers, librarians, newsagents and the like)[26] that mitigates the harshness of the strict publication rule must be contrasted with the common law principles derived from Bryne v Deane[27] concerning owner(s) or occupier(s) whose premises are defaced without their approval or encouragement by trespassers who have affixed defamatory statements onto a notice board or scrawled such statements as graffiti on the wall. Such owner(s) or occupier(s) have not, in any sense, assisted or participated in the publication of the libel and are not publishers of the libel. But failure to remove defamatory material from the property within the control of such owner(s) or occupier(s) can in some circumstances make them responsible for publication of the statement. Thus, the Bryne principle is “not about relieving a person’s prima facie liability as a publisher [ie the defence of innocent dissemination] but about whether or not an occupier is constituted a publisher”.[28]
60. In Bryne, the claimant was a member of a golf club whose owners illegally kept gambling machines on the premises. Someone reported to the police and afterwards a poem was posted in the club implying that the plaintiff was the informant. On appeal, it was held that the words were not capable of a defamatory meaning. Greer LJ at pp 829-830 said:
“…… the two defendants are lessees of the club and they are the occupiers of the club premises, and the walls are their walls, and in my judgment they allowed a defamatory statement to be put up on their walls and to remain on their walls in a position in which it could be read by anybody who came to the club. …… In my judgment the two proprietors of this establishment by allowing the defamatory statement, if it be defamatory, to rest upon their wall and not to remove it, with the knowledge that they must have had that by not removing it it would be read by people to whom it would convey such meaning as it had, were taking part in the publication of it.”
Greene LJ agreed and posited that the defendants should be liable if having regard to all the facts of the case the proper inference was that by not removing the defamatory statement they really made themselves responsible for its continued presence in the place where it had been put.[29] Such inference would ordinarily be drawn unless it would require very great trouble and expense to remove the defamatory material. Slesser LJ thought only the secretary had published the defamation by adopting the statement posted on the notice board.[30]
61. Greene LJ’s judgment was approved by Hunt J of the New South Wales Supreme Court in Urbanchich v Drummoyne Municipal Council.[31] That case concerns posters with photographs of persons in Nazi uniforms in the company of Adolf Hitler glued to bus shelters in the control of the Urban Transport Authority. The face of one of the persons was identified as the plaintiff, and the accompanying text allegedly contained defamatory imputations. The plaintiff drew the Authority’s attention to the posters and asked for their removal, but they remained in place for another month. Relying on Bryne, Hunt J held the Authority was capable of being a publisher if the jury found there was “an acceptance by the defendant of a responsibility for the continued publication of that statement”:
“In a case where the plaintiff seeks to make the defendant responsible for the publication of someone else’s defamatory statement which is physically attached to the defendant’s property, he must establish more than mere knowledge on the part of the defendant of the existence of that statement and the opportunity to remove it. According to the authorities, the plaintiff must establish that the defendant consented to, or approved of, or adopted, or promoted, or in some way ratified, the continued presence of that statement on his property so that persons other than the plaintiff may continue to read it ……”
This suggests that an inference of publication would only be drawn after the defendant had a reasonable period after the notification to take down the defamatory material. Hunt J emphasised that whether any inference of publication should be drawn is a question of fact. His reasoning seems to reject an imposition of a positive duty to remove defamatory material except when there is an assumption of responsibility for continued publication.
62. In Fevaworks Solutions Ltd, Ribeiro PJ concluded that the Bryne line of cases was authority for the following propositions:[32]
“(a) Where a third person writes or affixes a statement defamatory of the plaintiff on the occupier’s property without the occupier’s knowledge, the occupier is not treated as a publisher of that statement prior to his becoming aware of it.
(b) Once the occupier discovers its existence, he may be treated as a publisher but only if, having the power to do so, he does not remove or obliterate the offending statement in circumstances which justify inferring as a matter of fact that by his inaction he has consented to or ratified its continued publication.
(c) Where the occupier becomes aware of the libel but the circumstances show that removal or obliteration is very difficult or very expensive, the fact that the defamatory statement is not expunged may well not justify the inference that it remains in place with his approval.”
63. The distinction between the two sets of principles explained in paragraph 59 above underlies Ribeiro PJ’s disagreement with the treatment of the Bryne principle as a facet of the innocent dissemination defence in some of the reported cases involving libel in an internet context.[33] This fundamental distinction also underlies his conclusion that the Bryne principle does not apply to internet platform or discussion forum providers who cannot be equated as “the occupier of premises who is not in the business of publishing or facilitating publication at all, but who has imposed on him the defamatory act of a trespasser”. Since internet platform or discussion forum providers in that case played an active role in encouraging and facilitating the multitude of internet postings by their members, they were participants in the publication of postings by the forum’s users and they fell from the outset within the broad traditional concept of “a publisher”, a characteristic not shared by a golf club or other occupier who puts up a notice board on which a trespassing message is noted.[34]
64. On this basis, Ribeiro PJ parted company with the reasoning adopted in Tamiz v Google Inc[35] and Wishart v Murray & ors,[36] which reasoning proceeds on the basis that successful invocation of the defence of innocent dissemination results in the defendant being deemed not to have published at all. Rather, internet discussion platform providers (such as the defendants before him) that encourage and facilitate publication falls within the concept of “a publisher” under the strict publication rule, and it is for them to raise the defence of innocent dissemination (ie they have no actual knowledge of the alleged libel, are not aware of circumstances that give cause to suspect a libel, and are not negligent in failing to discover the libel) if they can.[37]
65. It is therefore clear from Fevaworks Solutions Ltd that the primary question is whether or not the defendant is within the broad traditional concept of a publisher under the strict publication rule: [38]
a) If he is found to be a non-publisher and the defamatory act of a trespasser was imposed on him, the relevant question under the Bryne principle is whether he has at some point been fixed with knowledge of the libel and has demonstrated his consent to and adoption of the libel, and has thus turned himself into a publisher.
b) If he is found to be a publisher who has participated in the publication, the relevant question is whether he is entitled to rely on, and have established, the defence of innocent dissemination in order to be relieved of the strict publication rule that is otherwise applicable.
To address this question, one has to ascertain whether he is a subordinate or main publisher since such defence is only available to the former and not the latter. But it must be remembered that “[the] starting point of the discussion is that a subordinate distributor is equally liable as the main publisher of the libel” (ie he is nonetheless a publisher), and “[it] is against the harshness of this rule that the innocent dissemination defence is made available to the subordinate distributor”.[39]
If the defendant knows the content of the defamatory material and has authorised/participated in its publication, he will be liable as a main publisher.[40] Ribeiro PJ also held that an internet intermediary is a main publisher where he knows or ought reasonably to be aware of the content of the article complained of though not necessarily of its defamatory nature as a matter of law (ie the knowledge criterion) and has a realistic ability to control publication of such content (ie the control criterion).[41]
66. On the facts of Fevaworks Solutions Ltd, the defendants were providers, administrators and managers of a website that hosted a popular internet discussion forum. Only registered members could post and respond to topics on the forum, but all posts could be read by the public on the website. Certain forum members posted anonymous messages which suggested the plaintiffs were involved in crimes and triad activities. The issues included whether the defendants were in law regarded as publishers of the defamatory postings and whether the common law defence of innocent dissemination was applicable. Ribiero PJ concluded that the defendants were subordinate publishers of the defamatory postings “since they provided the platform for their dissemination”, but until notified they neither knew nor was it reasonable to expect them to know of the libellous statements, hence they “are therefore in principle able to invoke the innocent dissemination defence as subordinate publishers”.[42] Once they were notified, they took all reasonable steps to remove the offending publication from circulation as soon as reasonably practicable, and they successfully relied on such defence.
67. In short, the defendants in Fevaworks Solutions Ltd have been found to be publishers under the strict publication rule (ie they intentionally assisted in providing a platform for the dissemination of the defamatory postings regardless of whether they knew such postings contained defamatory material), which is a separate issue from qualification for main or subordinate publisher.[43] In any event, they have satisfied the requisite mental element for the act of publication, which is not knowledge of the content of the article being published,[44] but rather an intention to “lend his assistance to [the existence of the defamatory material] for the purpose of being published”.[45] It is the defendant’s instrumentality to the publication rather than knowledge of defamatory content that is relevant.
68. I have discussed the aforesaid distinction at some detail because the case law from other common law jurisdictions must now be viewed through the prism of such distinction drawn in Fevaworks Solutions Ltd, which is binding on me.
69. The early internet case of Godfrey is an action in defamation against an ISP that carried a newsgroup that was available to be read by customers. An unknown person in the United States posted a comment that was considered defamatory of the plaintiff. Morland J held that the transmission of a defamatory posting from the storage of a news server constituted publication of that posting to any subscriber who accessed the newsgroup containing that posting. He found the defendant was not a mere passive owner of an electronic device through which postings were transmitted, but it chose to receive and store the newsgroup containing the offending postings on its computers and transmit them in response to requests. In other words, the defendant was not a mere conduit for it hosted and transmitted offending material, which it could delete if it chose.[46]
70. Morland J accepted that liability for publication at common law was strict and therefore the ISPs could be treated as publishers of defamatory material stored on their servers even though they were ignorant of their existence. He found they would be liable unless they could establish (and the onus was on them) the defence of, say, innocent dissemination.[47] Godfrey was cited without disapproval in Loutchansky v Times Newspaper Ltd & ors (Nos 2-5).[48] Morland J’s reasoning is in line with the law as explained in Fevaworks Solutions Ltd.
71. But the more recent English common law authorities appear to treat knowledge of the existence of the defamatory statement as essential before an internet intermediary can be considered to be a publisher.[49]
72. In Bunt v Tilley & ors,[50] Mr Bunt, who conducted his case in person, ran a website for his business which incorporated a discussion forum. He alleged that Mr Tilley (a) posted a “thread” on the website that disparaged his products and (b) with two others made defamatory comments about him on other discussion forums and newsgroups. Mr Bunt, relying heavily on Godfrey, sued their ISPs alleging they provided “corporate sponsorship and approval” of the defamatory postings and were the conduit used to publish the offending comments. Eady J held that Mr Bunt had no reasonable prospect of establishing any of the ISPs had participated in the relevant publications, and the ISPs successfully applied for summary judgment.
73. Eady J said that “…… to impose legal responsibility upon anyone under the common law for the publication of words it is essential to demonstrate a degree of awareness or at least an assumption of general responsibility”,[51] and it was necessary to demonstrate an ISP’s activity was more than passive and that it knowingly involved in the process of publishing “the relevant words”. Consequently, “persons who truly fulfill no more than the role of a passive medium for communication cannot be characterised as publishers: thus they do not need a defence”.[52]
74. It appears that the passivity rule (ie “mere conduits” or “passive facilitators” do not attract liability), if I may so describe it, in the ratio of Bunt is different from the obiter view of Morland J in Godfrey.[53] The question of knowledge as raised by Eady J in Bunt has been considered by Ribeiro J in Fevaworks Solutions Ltd in the context of the knowledge criterion for determining whether or not a publisher under the strict publication rule is a main publisher. The plaintiffs in that case argued that so long as the defendants knew they were hosting and making accessible a multitude of postings on the forum they must be taken to know the contents of the postings or discussion threads complained of since they formed part of that multitudinous body of material.[54] But Ribeiro PJ (following Eady J’s observations in Bunt) disagreed, and said the knowledge criterion for finding the publisher to be a main publisher “should be taken to mean that the publisher must know or be taken to know the content – not necessarily every single word posted – but the gist or substantive content of what is being published”, but it is “irrelevant whether the provider realised that such content was in law defamatory”.[55]
75. As explained above, such observations by Ribeiro PJ were made in the context of ascertaining whether or not a provider (being a publisher under the strict publication rule) “qualify as a first or main publisher”.[56] “It should be stressed that adopting a more focused requirement as to knowledge does not mean absolving a platform provider from liability. It means treating it as a subordinate publisher and throwing on it the burden of bringing itself within the innocent dissemination defence.”[57] This does not address the antecedent question of whether or not the provider qualifies as a publisher, and Bunt must be read in such context in this jurisdiction. Indeed, even Gatley suggests that “it is certainly arguable that, like others innocently involved in the transmission of a defamatory publication, [the defendants in Bunt] should have been treated as subordinate publishers. It has been said to be the law that the mere transmission by an intermediary is a publication”.[58]
76. As regards the control criterion for identifying the main publisher in contra-distinction to the subordinate publisher, Ribeiro PJ[59] cited Eady J’s reference to “an opportunity to prevent publication” in Bunt, and also to the matter of editorial control emphasised by the Supreme Court of Canada in Crookes v Newton.[60] At issue in the latter case was whether creating an internet hyperlink to defamatory material constituted “publication” for the purpose of defamation law.
77. In that case, after C brought defamation actions against various parties alleging he had been defamed in several articles on the internet, N posted an article on his website which commented on the implications of C’s defamation suits and included hyperlinks connected to the impugned articles. C asked N to remove the hyperlinks, but N refused. It was held that providing hyperlinks did not amount to publication of the linked article, and the majority emphasised that the author who created the hyperlink did not control the secondary article’s content or publication. N therefore had not published any defamatory material, and C’s action could not succeed.
78. The judgment considered the breath and harshness of strict publication rule, the amelioration brought by development of the defence of innocent dissemination defence, and the fact that in recent years the strict publication rule had been tempered by cases (eg Bunt and Metropolitan International Schools Ltd v Designtechnicia Corporation[61]) which suggest that some acts of communication are so passive that they should not be considered publication. In light of these developments, it was held that creating a hyperlink (which essentially is a content-neutral reference and which gives the author no control over the content of the linked secondary article) to defamatory material is not the type of act that constitutes publication. It was said that hyperlinks only communicate that something exists, but do not by themselves communicate content, especially when some further act by a third party is required before the content can be accessed.
79. As explained above, Riberio J’s discussion of Crookes was in the context of the control criterion for determining whether or not a provider qualifies as a main publisher (failing which the provider will be treated as a subordinate publisher), but Abella J, writing for the majority in Cookes, held that a formalistic application of the strict publication rule to hyperlinks was not appropriate and relied on the passivity rule (as explained above) to say that hyperlinks did not constitute publication of the linked material.
80. I bear in mind that Ribeiro PJ did “note in passing” the argument in certain internet cases whether internet intermediaries “should be regarded as “mere conduits” or “passive facilitators” comparable to the Post Office or a telephone company”, but there is no detailed discussion since it was not suggested that the defendants in Fevaworks Solutions Ltd were such.[62] In my view, any reliance on the passivity rule (as explained above) in this jurisdiction must be viewed in light of the ratio and approach in Fevaworks Solutions Ltd as discussed above, especially in relation to the strict publication rule for determining qualification as a publisher, which is a separate issue from categorising a publisher as a subordinate publisher for invoking the innocent dissemination defence. It is interesting to note that Gatley says the “better view” may be that internet intermediaries that “facilitate” access to websites provided by others (eg providing the computer systems through which communications happen to pass on their route from one computer to another) “should be treated as subordinate publishers in that they intentionally participate in the dissemination of the material complained of”,[63] and it is arguable that under the strict publication rule:[64]
“not only should the internet service providers in Bunt v Tilley have been treated as having published the defamatory statements but it would seem to follow that the Post Office “publishes” a defamatory letter which it delivers and a telephone company publishes a slander when it relays a defamatory telephone call. Of course, in the overwhelming majority of cases internet service providers, the Post Office and telephone companies do not know, nor is it reasonable to expect them to know, of the contents of the publications and therefore they will be able to rely on either the common law defence of innocent dissemination or one of the statutory defences [which is not applicable to Hong Kong].”
81. In any event, Crookes expressly left open the question whether the same principles (ie it is not publication of the referenced content absent evidence of someone having actually viewed and understood the defamatory information to which the user-activicated hyperlink directs) would apply to embedded or automatic links that automatically display referenced material with little or no prompting from the reader.[65] According to the majority’s reasoning, if hyperlinks are made to appear as part of the website that the hyperlinker controls (ie the use of the hyperlink in itself conveys a defamatory meaning about the plaintiff as where the link repeats the defamatory sting conveyed by the underlying content), it can arguably constitute publication.
82. In Davison v Habeeb,[66] HH Judge Parkes QC set aside the order of a master granting permission to serve the writ on Google Inc in the United States. That case concerned allegations published in an online newspaper about the claimant, and later republished by the author and another on his personal website managed by Google Inc’s blogging platform “”. The claimant asked Google Inc to remove the material, but Google Inc declined.
83. In that case, the judge found to be like a gigantic notice board, and concluded that Google Inc had the ability to “take the notices down …… if they are pointed out to it” (ie liability was dependent on receiving notification that it was carrying defamatory material).[67] Whilst the judge expressed sympathy for Google Inc’s position,[68] he concluded that “[in] my view it must be at least arguable that [Google Inc] should properly be seen as a publisher responding to requests for downloads like Demon Internet, rather than a mere facilitator, playing a passive instrumental role”.[69] Since Google Inc was a publisher at common law, following notification it would be unable (or at least arguably unable) to establish it was ignorant of the defamatory material on , and at some point after notification it would become liable for continued publication of the material complained of on the Bryne principle of consent or acquiescence.[70]
84. Even though HH Judge Parkes QC held that Google Inc was a publisher of the material complained of and at least following notification they would be liable for publication of that material, he concluded there was no realistic prospect of the claimant establishing that notification of the complaint fixed Google Inc with actual knowledge of unlawful activity or information in that case since Google Inc was not in the position to adjudicate on the conflicting claims from the claimant and the 2nd defendant.[71] Had the complaint been “sufficiently precise and well substantiated, and where there was no attempt by the author of the defamatory material to defend what had been written”, a different conclusion might have been reached.[72]
85. But Ribeiro PJ has already rejected the approach adopted in Davison which conflated the Bryne principle and the defence of innocent dissemination by treating the former merely as a facet of the latter.[73] Since HH Judge Parkes QC held in Davison that it was at least arguable Google Inc was more than a mere facilitator playing a passive instrumental role and could be seen as a provider hosting material on its servers and responding to requests for downloads, it must also be arguable, following Ribeiro PJ’s reasoning, that it would have been found to be a publisher under the strict publication rule in this jurisdiction.
86. I now turn to Tamiz v Google Inc[74] and its reasoning which Ribeiro PJ expressly declined to adopt. In that case, the claimant sued in respect of defamatory comments posted on a blog hosted by that was operated by Google Inc. The claimant complained to Google Inc about the comments and after some delay Google Inc contacted the author of the blog who voluntarily removed it. The claim concerned comments published after Google Inc was notified of the comments but prior to their removal. Eady J set aside service of the claim form out of jurisdiction, and in reaching that decision he found Google Inc was not the publisher of three of the comments which were arguably defamatory, and the period between notification and removal was so short and any potential liability so trivial that it was Jameel abuse of process to maintain the proceedings. Mr McCoy SC reminded that Eady J took note it was virtually impossible for Google Inc to exercise editorial control over the content of the blogs it hosted which in the aggregate contained more than half a trillion words, with 250,000 new words added every minute.[75]
87. In the English Court of Appeal, Richard LJ felt it unnecessary to address counsel’s criticisms of Bunt, but he was not persuaded that Eady J fell into any fundamental error of analysis or reached the wrong conclusion in relation to the kind of internet service under consideration in that case.[76] However, he held that in the case before him Google Inc was neither a primary nor secondary publisher prior to notification of complaint,[77] but they fell within the scope of Bryne v Deane[78] after notification. Richards LJ said at p 321 that:
“ 33. …… The provision of a platform for blogs is equivalent to the provision of a notice board; and Google Inc goes further than this by providing tools to help a blogger design the layout of his part of the notice board and by providing a service that enables a blogger to display advertisements alongside the notices on his part of the notice board. Most importantly, it makes the notice board available to bloggers on terms of its own choice and it can readily remove or block access to any notice that does not comply with those terms.
34. Those features bring the case in my view within the scope of the reasoning in Bryne v Deane. Thus, if Google Inc allows defamatory material to remain on a Blogger blog after it has been notified of the presence of that material, it might be inferred to have associated with, or to have made itself responsible for, the continued presence of that material on the blog and thereby to have become a publisher of that material. ……
36. The period during which Google Inc might fall to be treated on that basis as a publisher of the defamatory comments would be a very short one, but that means that the claim cannot in my view be dismissed on the ground that Google Inc was clearly not a publisher of the comments at all.” (my emphasis)
88. The English Court of Appeal plainly found that Google Inc facilitated publication of the blog and its comments, hence the effect of Tamiz is to sweep away the strict publication rule under common law “certainly so far as concerns internet publications” in the English jurisdiction, which view has been rejected by the Court of Final Appeal in Fevaworks Solutions Ltd which treated a website operator/host as a subordinate publisher who made facilities available for others to post and thereby intentionally participated in the dissemination of the defamatory material, and such website operator/host was liable unless it could bring itself within the defence of innocent dissemination which required it to show that it neither knew nor ought to have been aware of the content of the material.[79]
89. Mr McCoy SC urges me to apply the passivity rule (as explained above) as developed in Bunt, and says that the facts of Tamiz (which concerned a blog platform host who would more fairly be expected to administer the content of the postings by internet users) are crucially different from those in the present case (which concerns a search engine that passively reproduces content placed freely on the internet by internet users). On the other hand, whilst Mr Ng agrees that Tamiz was not concerned with internet search engine operator, he submits Bunt must now be reviewed in light of Fevaworks Solutions Ltd which has diverged from the English authorities.[80] He says that consequently Bunt is now irrelevant to the primary question of whether or not Google Inc as a search engine is a “publisher” of the Words on the Websites.
90. In Wishart v Murray & ors,[81] the defendants set up a Facebook page to campaign for a newly published book to be boycotted, and they made defamatory comments against one of the authors. One of the defendants monitored and moderated postings on the Facebook page that was created. Courtney J considered the Byrne principle (followed in Urbanchich) and also other English authorities (eg Godfrey, Bunt, Metropolitan International Schools Ltd, Davison and Tamiz), and found that a host of a Facebook page had in essence established a notice board which gave him the power to control the content published, and since the defendants could not be passive hosts of content posted on their Facebook page, they must be regarded as publishers of the postings made by anonymous users (a) if they knew of the defamatory postings and failed to remove them within a reasonable amount of time or (b) if they were not aware of the defamatory postings but based on the circumstances they ought to know that the postings made would be defamatory in nature. In short, Courtney J found that either actual or inferred knowledge of the defamatory imputations was sufficient to find liability as publisher for a host of a Facebook page in respect of comments by a third party.[82]
91. Mr McCoy SC submits this case is not helpful and is quite distinguishable as it deals with anonymous defamatory postings on a Facebook page created by the host, which bears no rational analogy with a search engine operator as it is possible and indeed practicable for the creator of a Facebook page to manage the content posted on it. There is no need for me to consider such submission because Wishart suffers from a more fundamental concern in this jurisdiction. Its reasoning has been rejected by Ribeiro PJ in Fevaworks Solutions Ltd on the same basis as for his divergence with Tamiz.[83] This has been recognised in Gatley:[84]
“As the Hong Kong Final Court of Appeal explained in Oriental Press Group Ltd v Fevaworks Solutions Ltd, Bryne v Deane applies where the occupier of premises has defamatory material forced on him: it does not apply where the defendant actively encourages third parties to post material on the website that it provides. The better view may therefore be that hosts are secondary publishers who will, in the absence of knowledge or imputed knowledge, be able to rely on the innocent dissemination defence.”
92. This brings me to the authorities in the common law jurisdictions on the liability of internet search operators. This was considered in yet another decision by Eady J in Metropolitan International Schools Ltd. In that case, the claimant sued in respect of defamatory bulletin board postings on a website. The libels appeared as snippets of information (accompanied by hyperlink) when the users carried out internet search under the claimant’s name using Google Inc’s search engine. Eady J found the internet search was performed automatically by the search engine’s computer programme, and involved no input by any officer or employee of Google Inc. He distinguished this from the compiler of a conventional library catalogue, who might be liable for compiling the catalogue on the basis that liability attached for repeating a libel. He also found Google Inc played no role in formulating the search
terms and could not prevent the snippets from appearing in response to the user’s search request unless some positive steps had been taken in advance, so the essential mental element required for establishing liability for publication was absent, and Google Inc could not be characterised as a publisher at common law but was merely a facilitator. Eady J therefore set aside the service out of jurisdiction.
93. The next question in that case was whether the legal position became different once Google Inc was informed of the defamatory content of a snippet thrown up by the search engine. Eady J held that although a person could become liable for the publication of a libel by acquiescence (ie by permitting publication to continue when he/she has the power to prevent it),[85] a search engine was different from an internet intermediary or website host in that it had no control over the search terms to be typed by future users, and if a future search result throws up the impugned words, it does not follow that Google Inc has authorised or acquiesced in that process.[86] Whilst Google Inc was able to (and eventually did) prevent searches returning links to specific URLs (uniform resources locators) which had been identified to them by the claimant, it was not technically able to put in place a more effective block on the impugned words without at the same time disabling access to a huge amount of other material. So even if Google Inc’s “notice and take down procedure” had not operated as rapidly as it should, they did not become liable as a publisher of the offending material, whether on the basis of authorisation, approval or acquiescence, in the period between notification and removal.[87]
94. In Fevaworks Solutions Ltd, Ribeiro PJ discussed Metropolitan International Schools Ltd in the context of the defence of innocent dissemination, which requires the defendant to prove that he did not know and would not, with the exercise of reasonable care in the relevant circumstances, have known that the article contained defamatory content. In relation to subsequently acquired knowledge, the internet intermediary may satisfy the requirements of the defence in terms of lack of knowledge etc, but “may subsequently become aware of the relevant defamatory content in circumstances where the defamatory article is still available for distribution and where, before it can removed, some further acts for publication in fact occur”. However, Ribeiro PJ said:
“94. The existing authorities have tended to deal with this situation under the rubric of the “notice board” cases as discussed above. But as previously stated, I do not consider that analysis satisfactory. The issue appears to me to be better viewed as arising in the context of the innocent dissemination defence.
……
96. In the situation posited, the premise is that the defendant is a subordinate publisher who made the article available when there was nothing in the work or the circumstances under which it came to him or was disseminated by him which should have led him to suppose that it contained a libel; and his lack of awareness was not due to any negligence on his part. The defence implicitly recognised that such a defendant may in fact unwittingly publish matter that is defamatory of the plaintiff, but exempts him from liability provided he meets the standard of reasonableness.” (my emphasis)
It was in such context that Ribeiro PJ referred to Eady J’s observations in Metropolitan International Schools Ltd concerning the position after notification was made to the search engine operator.
95. Eady J in Metropolitan International Schools Ltd held that under English law an automated search engine operator is not liable as publisher in respect of search results appearing as snippets as it never had control over the result of a search and lacked any knowing involvement in publication, but if, as explained by Ribeiro PJ, the question of knowledge of content is to be brought within considerations for the innocent dissemination defence, it begs the antecedent question as to whether the search engine operator may “unwittingly” be a publisher under the strict publication rule leaving it to raise the defence of innocent dissemination (if it can). Eady J appeared to offer a negative answer to this question by characterising Google Inc as a mere “facilitator” for communicating others’ content without human involvement. But such conclusion is doubtful in this jurisdiction in view of Ribiero PJ’s reasoning in Fevaworks Solutions Ltd.
96. Indeed, other parts of the Commonwealth jurisdictions have taken a different approach. But before I turn to such cases, I should briefly mention Budu v the British Broadcasting Corporation,[88] which deals with the reverse situation to Metropolitan International Schools Ltd. One of the issues in Budu was whether the originator of the defamatory material would be liable for the publication of a search engine snippet. In that case, the claimant brought libel proceedings against the defendant for three articles on its website. He also sued the defendant for “republication” in respect of a snippet extracted from the second article which appeared with the hyperlink to it when a “google” search was done on the claimant’s name. It was held that the second article did not bear the meaning alleged by the claimant, so there was no “republication”. In coming to such conclusion, Sharp J followed Metropolitan International Schools Ltd (ie the service engine operator could not be held liable for the snippet), but rejected the claimant’s argument that it must therefore follow that the defendant would be liable. He concluded that liability for publication could not accrue by default,[89] but observed (following Metropolitan International Schools Ltd) that a search engine provider “might at some point be liable, if the publication of a defamatory search result (or a “signpost to a conduit” ……) continued after notification of the specific URL from which the words complained of originated ……”[90] Again, this must also be read in light of the above observations on Metropolitan International Schools Ltd with relevance to this jurisdiction.
97. As said, a different approach was taken in other common law jurisdictions. In Trkulja v Yahoo! Inc LLC & anor,[91] the plaintiff claimed that a “yahoo” search using his name returned results (article and images) that associated him with organised crime and criminal figures, which he claimed was defamatory. The defendants by their defences pleaded that subject to the plaintiff establishing at trial that any person had downloaded and read the matter using the Yahoo! 7 internet search engine, they admitted they published the matter to such person. The plaintiff called two witnesses at trial to give evidence that they had downloaded and read the article using the defendants’ internet search service. The defendants therefore conceded they had published the articles to such persons.
98. In Trkulja v Google Inc LLC & anor (No 5),[92] the same plaintiff in Trkulja v Yahoo! Inc LLC & anor claimed that a “google” search using his name returned similar results as the “yahoo” search. He complained to Google Inc, but they declined to take down the results. The jury found Google Inc was a publisher but an innocent disseminator until they received notification. Having concluded that Google Inc had knowledge of the defamatory material after notification, the jury awarded the plaintiff AUD200,000 in damages. Notwithstanding the jury’s verdict, Google Inc applied for judgment on the argument that as a matter of law[93] it was not a publisher since it had not “consented to, or approved of, or adopted, or promoted, or in some way ratified, the communication of the material complained of”.[94] However, Beach J distinguished the English authorities (including Metropolitan International Schools Ltd), and concluded that since Google Inc’s algorithmic processes operated and generated content precisely as they intended, the requirement that an entity intentionally lent its assistance to publication in order for such entity to be held responsible for the defamatory material was satisfied, and on such basis he rejected the characterisation of Google Inc as an “internet intermediary” and held there was basis for the jury to conclude they were a publisher for the purpose of defamation law.
99. Since the judgment of Beach J touched on an essential argument by Google Inc and Mr McCoy SC in the present case, ie the automated nature of Google Inc’s algorithmic systems and the lack of human input in Google Search’s search processes, I propose to set out his judgment in some detail:
“16. The plaintiff accepted (correctly in my view) that he had to establish that Google Inc intended to publish the material complained of. While much was made by counsel for Google Inc of the fact that there was no human intervention between the request made to the search engine and the publication of search results, and of the fact that the system was ‘fully automated’, the plaintiff’s point was that Google Inc intended to publish everything Google’s automated systems (which systems its employees created and allowed to operate) produced. Specifically, the plaintiff contended that Google Inc intended to publish the material complained of because while the systems were automated, those systems were the consequence of computer programs, written by human beings, which programs were doing exactly what Google Inc and its employees intended and required. On this basis, it was contended that each time the material complained of was downloaded and comprehended, there was a publication by Google Inc (the operator and owner of the relevant search engines), as intended by it. So it was submitted by the plaintiff that Google Inc was a publisher throughout the period in respect of which complaint was made.
……
18. …… it was open to the jury to find the facts in this proceeding in such a way as to entitle the jury to conclude that Google Inc was a publisher even before it had any notice from anybody acting on behalf of the plaintiff. The jury were entitled to conclude that Google Inc intended to publish the material that its automated systems produced, because that was what they were designed to do upon a search request being typed into one of Google Inc’s search products. In that sense, Google Inc is like the newsagent that sells a newspaper containing a defamatory article. While there might be no specific intention to publish defamatory material, there is a relevant intention by the newsagent to publish the newspaper for the purposes of the law of defamation.
19. By parity of reasoning, those who operate libraries have sometimes been held to be publishers for the purposes of defamation law. That said, newsagents, librarians and the like usually avoid liability for defamation because of their ability to avail themselves of the defence of innocent dissemination (a defence which Google Inc was able to avail itself of for publications of the images matter prior to 11 October 2009, and all of the publications of the web matter that were the subject of this proceeding).
20. As was pointed out by counsel for the plaintiff in his address to the jury, the first page of the images matter ….. was a page not published by any person other than Google Inc. It was a page of Google Inc’s creation – put together as a result of the Google Inc search engine working as it was intended to work by those who wrote the relevant computer programs. It was a cut and paste creation (if somewhat more sophisticated than one involving cutting word or phrases from a newspaper and gluing them onto a piece of paper). If Google Inc’s submission was to be accepted then, while this page might on one view be the natural and probable consequence of the material published on the source page from which it is derived, there would be no actual original publisher of this page.
……
27. As the judgments in Bunt, Metropolitan Schools and Tamiz show, and as Eady J acknowledged in Tamiz, the question of whether a particular internet service provider might be a publisher in respect of defamatory material published through or via or with the assistance of a particular internet product is ‘fact sensitive’. The facts in Bunt and Tamiz are substantially different from the facts in the present case – such that the conclusions in those cases give no real assistance to the resolution of the issue in the present case. The facts in Metropolitan Schools (involving a search engine) are more applicable to the present case. That said, it is not possible for me to say whether all of the evidence given in the present case was also presented to Eady J in the interlocutory application that his Lordship was required to determine. Specifically, his Lordship does not appear to have given any consideration to the fact that internet search engines, while operating in an automated fashion from the moment a request is typed into them, operate precisely as intended by those who own them and who provide their services. Additionally, his Lordship appears to have been moved to come to his conclusion in part because of the steps taken in that case by the relevant defendant (Google Inc) to block the identified URLs. This was a matter which, on the facts in the present case, was not capable of bearing upon the issue of whether there was publication by Google Inc as contended for by the plaintiff.
28. While much was made by Google Inc in the present case of Eady J’s statements in Bunt and Tamiz that an internet service provider who performs no more than a passive role cannot be a publisher, those statements have to be seen in the light of the facts in those cases. To say as a general principle that if an entity’s role is a passive one then it cannot be a publisher, would cut across principles which have formed the basis for liability in the newsagent/library type cases and also in those cases where someone with power to remove a defamatory publication chooses not to do so in circumstances where an inference of consent can be drawn.
29. …… Further, while on the facts in Bunt, the defendants were correctly described as ‘internet intermediaries’ (whatever may be the legal consequences of such a description), it is, with respect, doubtful that that same description can be applied to an internet search engine provider in respect of material produced as a result of the operation of that search engine. That said, any such ‘internet intermediary’ is, in any event, performing more than the ‘merely passive role ... [of] facilitating postings’ (Cf Bunt).
30. …… Google Inc goes further and asserts that even with notice, it is not capable of being liable as a publisher ‘because no proper inference about Google Inc adopting or accepting responsibility complained of can ever be drawn from Google Inc’s conduct in operating a search engine’.
31. This submission must also be rejected. The question is whether, after relevant notice, the failure of an entity with the power to stop publication and which fails to stop publication after a reasonable time, is capable of leading to an inference that that entity consents to the publication. Such an inference is clearly capable of being drawn in the right circumstances (including the circumstances of this case). Further, if that inference is drawn then the trier of fact is entitled (but not bound) to conclude that the relevant entity is a publisher. Google Inc’s submission on this issue must be rejected for a number of reasons, the least of which is that it understates the ways in which a person may be held liable as a publisher.”
100. Mr McCoy informs me there is a pending appeal against Beach J’s judgment. He also asks me to approach this case with caution because the judgment was concerned with an application to overturn a jury verdict in favour of the plaintiff, and hence Beach J’s focus was merely to consider whether the jury verdict was reasonably open to a properly instructed jury. Mr McCoy SC submits that in any event Beach J has put the proposition too broadly, and he urges me to uphold the passivity principle in the English cases. He also attempts to distinguish Trkulja v Google Inc LLC on the basis that it concerned images rather than autocompletions, and he reminds that Yeung’s claim is a novel one unsupported by the earlier authorities.
101. But Trkulja v Google Inc LLC is not a lone authority. It was followed by A v Google New Zealand Ltd[95] in New Zealand. In that case, defamatory statements were posted on a website in the United States about the plaintiff, so when his name was “googled” the search results included a hyperlink to the United States website with snippets extracted from such website which contained defamatory material. The plaintiff asked the defendant to block the search results. This was done on occasion, but it failed to prevent the information and hyperlinks from being republished. The defendant applied for summary judgment and to strike out the claim on various grounds, including the argument that a search engine provider was not responsible under the law of defamation for the words appearing in automated search results. The court reviewed the earlier authorities, and concluded as follows:
“[68] Whether or not search engines are ‘publishers’ is a novel issue in New Zealand. It will be apparent from the above that its resolution requires determination of complex issues of law in a proper factual context. There may be need to consider whether there is ‘ a stamp of human intervention’ (as counsel for the plaintiff puts it) in the way that the search engine programme is written, and to address public policy concerns as occurred in Crookes v Newton.
[71] In light of this being a novel point of law and the limited foreign authorities available, it is reasonably arguable that a search engine is a publisher in respect of specific URLs and words and that, given the lack of clarity surrounding the subject of the plaintiff’s complaint, it is inappropriate to strike out that claim.”
102. Mr Ng says that Goggle Inc ran similar arguments as in Trkulja v Google Inc LLC in the present case, but in light of the reasoning of Trkulja v Google Inc LLC and A v Google New Zealand Ltd, which is in line with that of Ribeiro PJ in Fevaworks Solutions Ltd, and further bearing in mind that in the present case (a) Google Inc has conceded that they provided the search services upon accessing the Websites and (b) the Words were published on such Websites owned and operated by Google Inc (and not on any third party platform), Mr Ng urges me to follow the Australian and New Zealand jurisprudence which, he says, has overtaken the English cases and are consistent with Fevaworks Solutions Ltd. Mr Ng submits that at the very least there must be a good arguable case that Google Inc is a publisher of the Words on the Websites.
103. In my view, as a matter of law, it is plainly arguable that a search engine (including Google Inc) that generates objectively defamatory materials by its automated processes is a “publisher” within the meaning as explained in Fevaworks Solutions Ltd. It is also arguable that Google Inc by creating and operating its automated systems that generate materials in the manner they and their staff intended satisfies the requisite mental element for the act of publication, ie they provided the platform for dissemination and/or encouraged/facilitated or actively participated in the publication with intent to assist in the process of conveying the impugned words to publishees, and knowledge of defamatory content is not necessary (especially for an unwitting publisher who may in due course wish to invoke the defence of innocent dissemination if it can). I have not forgotten that unlike the non-publisher in Bunt Google Inc here is in the business of disseminating information, and it is arguable that they fall within the broad traditional concept of “a publisher”.
104. Indeed, this can be said to be a good arguable case given (a) the distinction drawn in Fevaworks Solutions Ltd between the identification of a “publisher” which is strict (but subject to the defence of innocent dissemination if applicable) and the Bryne principle for turning a non-publisher into a publisher upon knowledge and adoption of the libel, and (b) the Court of Final Appeal’s departure from the approach in Tamiz and other overseas authorities that conflated the two concepts. I also bear in mind Beach J’s suggestion that the English Court of Appeal in Metropolitan International Schools Ltd did not have the benefit of the argument made to him that internet search engines, while operating in an automated fashion, did so precisely as intended by those who own them and who provide their services. To this, I add that the English case law must now be viewed through the prism of the reasoning in Fevaworks Solutions Ltd which lends support to such argument.
105. Mr McCoy SC places emphasises on the passivity rule (as explained above) in Bunt and Metropolitan International Schools Ltd as well as Eady J’s observations in Tamiz to say that an internet search engine who performs a mere passive role cannot be a publisher. In my view, Beach J’s countervailing observation that if this is put forward as a general principle it cuts across the orthodox publication rule under which liability is strict also has arguable merits. It is also plainly arguable that under the common law principles of publication that one can unwittingly become a publisher, and then one may (if one can) invoke the defence of innocent dissemination.
106. I agree with Mr Ng that the issue of whether an internet search engine is a “publisher” in respect of Autocomplete and Related Searches suggestions is a controversial question in a developing area of law with limited foreign authorities and complex interplay between such foreign authorities and the principles found in Fevaworks Solutions Ltd. As explained above, the issue of whether a person or entity is a “publisher” is a matter of mixed fact and law. Indeed, Beach J said this is fact sensitive, and A v Google New Zealand Ltd also agreed that the issue of law must be resolved in a proper factual context. In my view, the issue of law should not be resolved summarily, and the proper factual context should be elicited for determination of the law on actual and not hypothetical basis.[96] Ma CJHC (as he then was) in Noble Power Investments Ltd v Nessei Stomach Tokyo Co Ltd explained:[97]
“18. Often (and the present case provides such an example), there will be a dispute over whether a claim does in fact come within one of the sub-heads of O.11, r.1(1). In such a situation, the court must be satisfied there is a good arguable case that the claim does fall within one of the requisite sub-heads: see Hong Kong Civil Procedure 2008, Vol.1 at pp.104-105 para.11/1/8. This test is applicable when there are disputes of fact or issues of mixed law and fact. Where the dispute is a succinct one of law, this ought to be resolved by the court even at the O.11 stage : see EF Hutton & Co (London) Ltd v Mofarrij [1989] 1 WLR 488 at p.495. I stress the word “succinct” as it may be inappropriate to go into complex questions of law at the O.11 stage.” (my emphasis)
107. As a matter of interest (although I need not rely on this authority), Collins on Defamation (2014)[98] cites the case of Rana v Google Australia Pty Ltd & ors[99] decided shortly after A v Google New Zealand Limited but before the hearing of the Summons. In Rana, the applicant complained of defamatory material authored by the 2nd and 3rd respondents on websites hosted by Google Inc which also appeared as search results in and .au. The applicant applied for leave to serve Google Inc outside Australia, and for such purpose Mansfield J had to consider whether Google Inc could be considered the publisher of the information. He discussed the English and Australian authorities and concluded as follows:
“50. I do not consider that it is appropriate on a “service overseas” application to decide that Google Inc is not the publisher of the defamatory material. Put shortly, the law is this regard is not settled: Duffy v Google Inc [2011] SADC 178, [29] – [31].
……
58. In short, whether or not a search engine could be considered a publisher of defamatory material is not settled in Australia, and accordingly, it would not be appropriate to refuse to give leave to serve the proceedings and on that basis effectively dismiss Mr Rana’s claim against Google Inc: A v Google New Zealand Ltd [2012] NZHC 2352, [69].”
108. On this note, I turn the facts presently before me in the affidavit evidence to ascertain the relevant factual context to see if a good arguable case can be established that Google Inc is a publisher in respect of the Words on the Websites.
(b) Facts
109. Here, the focus is on Google Inc’s instrumentality to the publication of the Words on the Websites. The core of Mr McCoy SC’s submissions is that Google Inc adopts an algorithmic based approach that requires no human input, operation and/or manipulation in the search processes for the results to appear, and as such Google Inc is a mere passive facilitator within a legal safe harbour for avoiding liability for defamatory suggestions. Mr McCoy SC says Google Inc cannot police the huge volume of webpages on the internet and is not under a duty to do so in respect of the web it crawls. It is said that Google Inc has no control over the search terms entered by the users or over the material that is placed on the web that are linked to by Google Search as a result of searches made by the users. It is said that Google Inc does not create or host web content, and the automatic nature of the way in which the search results (which are more akin to hyperlinks as discussed in Crookes) are displayed exclude any intent to publish with knowledge, ie to defame.
110. Mr McCoy SC submits that Google Search operates by using algorithms to simply present webpages to the users without any human assessment of the correctness, credibility or reliability of the information presented, and Google Search’s users understand that Google Inc is not making any representations to them about whether the webpage content is truthful. Mr McCoy SC argues that as a matter of logic, the conjunction of keywords in the presentation of predictions and related terms displayed by the Autocomplete and Related Searches features are not meaningful statements or words/thoughts attributable to Google Inc, especially when they are not even aware of such predictions or search results that appear each time a search is conducted.
111. In light of this expressed lack of control, Mr McCoy SC contends that the Autocomplete and Related Searches functions are neutral tools, and their results do not attribute negative connotations to any individual or entity. It is suggested that both “Albert Yeung Triad” and “Police Commissioner Triad” as autocompletions do not impute triad connections and/or defamatory views. The predictions and related search terms are just a “snapshot” reflection of data relating to the queries previous users have input (for any number of reasons) to find web content as well as the content of the webpages indexed by Google Search, and it is for the users to make of it as they wish. This, it is said, distinguishes such functionalities from internet discussion blogs/forums where members can freely post comments that may be defamatory.
112. Mr McCoy SC submits that since (a) the dynamic algorithmic processes merely generate preliminary indicative-only predictions and/or suggestions for search (which is final objective) and not the ultimate content of the search results, (b) the variability of the resulting suggestions is responsive to changes in statistical interest in the subject matter on a worldwide basis and (c) the ultimate content of the search may either confirm or dispel rumours, the Autocomplete and Related Searches suggestions cannot constitute dissemination of information. In the example cited in the above paragraph, the eventual search results may confirm or deny triad connections, or otherwise reveal bad character or show exemplary character.
113. Mr McCoy SC argues that since the alleged defamatory activity is created solely by automatic and independent algorithmic processes that (a) draw from the universe of previous users’ searches and also from worldwide web content (with human input limited to the implementation/maintenance of the algorithms themselves), (b) interpret “google” search activity, and (c) ensure the Autocomplete and Related Searches functions turn up the most relevant or at the every least the most commonly or frequently searched results, it would be impossible for Google Inc to manually interfere with or monitor the search processes given the billions of searches conducted by Google Search and the mass of information Googlebot “crawls” over each day. Mr McCoy SC submits that the entire basis of the internet will be compromised if search engines are required to audit what can be assessed by users using their search tools. Such infinite duty is a much greater task than removal of postings from an internet discussion forum, and it should not be foisted on Google Inc. He says this is what distinguishes Google Search from an internet blog/forum that comprises a closed group of registered members who agree to abide by forum rules and where the owner has the capacity and/or responsibility for editorial control over content.
114. Mr Ng, on the other hand, focuses on the issue of notification and removal, and complains that Google Inc has not responded to the 1st and 2nd Letters, has failed to remove the Words from the Websites, and has continued to publish or cause to be published the Words during the Period when they clearly knew and/or ought to have known of the presence of the Words on the Websites.
115. In my view, on the factual aspects as evident from the affidavit evidence, the analysis of Google Inc’s liability for Autocomplete and Related Searches results is concerned with the question whether they are an information provider with a neutral tool or whether they act beyond the scope of simply making information publicly available.
116. Unlike the writer of a letter, Google Search does not display the predictions or suggestions first-hand. Search autocompletions are reconstituted aggregations based on what other users at one time have typed as search queries into Google Search. By its algorithmic processes, Google Search recombines and aggregates data from web content and previous typed search queries, and transforms such data into Autocomplete and Related Searches suggestions by focusing on multiple factors set by Google Inc through the algorithms they designed, revised and improved. Google Inc acknowledges that Google Search’s algorithms use ranking factors to organise or “rank” all searchable content available from the content providers to determine which websites are the best sources of information for the search request made by the user. They also employ dedicated engineers to evaluate search quality on a full-time basis and to constantly finesse their algorithms to improve the search process. In short, Googlebot is constantly being reviewed, revised and improved by Gooogle Inc. Based on such experimentation, evaluation and analysis, in 2010 Google Inc launched 516 improvements to search.
117. In this sense, whilst Google Search’s algorithms weigh several factors that may be outside its control (eg information posted on the internet, search queries made by other users and/or search selections the users may ultimately pick) to generate the suggestions, its algorithms are synthesising and reconstituting input query data by previous users and web content uploaded by internet users before publishing them as suggestions. It is therefore arguable that Google Search does not simply convey information, but its Autocomplete and Related Searches features act by providing information distilled pursuant to artificial intelligence set up by Google Inc themselves by virtue of the algorithms they have created and maintained to actively facilitate the search processes. This raises a question as to whether or not Google Inc is a neutral tool, and whether or not the Autocomplete or Related Search suggestions are merely machine-generated and non-meaningful jumble of keywords. Notwithstanding Mr McCoy SC’s persuasion, I bear in mind that for the purpose of the Summons, Google Inc does not take issue that the Words are defamatory of Yeung.[100]
118. It is also evident on the affirmation evidence before me that Google Inc is capable of censoring material. Google Inc acknowledges that on occasion it manually edits results “to improve the user experience”. To tackle security concerns, it may remove “pages from our index (including pages with credit card numbers and other personal information that can compromise security) ……” It may also intervene in the search results for legal reasons, eg to remove “child sexual-abuse content …… or copyright infringing material (when notified through valid legal process ……” It can also make manual exceptions when their algorithms falsely identify sites, ie “mistakenly [catch] websites” that should not have been identified. It also takes automated and manual actions to remove spam, eg attempts to “game their way to the top of the results” by a technique known as “keyword stuffing”. In light of the above, there may be room for factual debate as to what further blocking steps, if any, would be open for Google Inc to take or how effective they may be.
119. It has been suggested that a key consideration is that automation means Google Inc is not even aware of the predictions or search results generated let alone any analysis of content, which therefore excludes any intent to publish with knowledge, ie to defame. But under the strict publication rule, the requisite mental element is not knowledge of the defamatory content or any intent to defame, but rather whether the defendant has actively facilitated or intentionally assisted in the process of conveying the material bearing the defamatory meaning to a third party, regardless of whether he knew that the material in question is defamatory.
120. This brings the arguments back to a more fundamental question (when shed of the modernity and complexity of the internet): as a matter of general tort principle, should or should not a person/entity remain responsible in law for acts done by his/her tool, and what are the limits of such liability (if any)? This question underlies the reasoning in Trkulja v Google Inc LLC, ie there is material to go to the jury that Google Inc should not be able to escape liability through the use of automated systems being their tool for carrying out their intended purpose, and whether they can raise any innocent dissemination defence is a separate matter.
121. Neither counsel has cited any other authority apart from Trkulja v Google Inc LLC. But in my view, given the reasoning in Trkulja v Google Inc LLC and in Fevaworks Solutions Ltd on the approach for identifying “a publisher”, there is plainly a good arguable case that Google Inc is more than a passive facilitator vis-à-vis their Autocomplete and Related Searches features.
122. Solely as a matter of interest (but there is no need for me to rely on this authority), I note that just before Trkulja v Google Inc LLC was decided, the Supreme Court of British Columbia in Century 21 Canada Limited Partnership v Rogers Communications Inc & anor doing business as Zoocassa Inc[101] made some relevant observations. In that case, Century 21 and two individual real estate brokers sued Zoocasa and its parent company Rogers due to Zoocasa’s indexing and “webcrawling” of the website operated by Century 21 featuring its brokers’ property listing. Zoocasa operated a “vertical” search engine that indexed property listings from a number of real estate websites on its own website, and in doing so copied property descriptions and photographs from the Century 21 website for use on its own website, which act violated the Terms of Use of Century 21’s website. The court found Zoocasa liable for copyright infringement and breach of contract, but Rogers was not liable for Zoocasa’s actions. The court found Zoocasa had agreed to the Terms of Use when it employed a search engine to crawl and fetch portions of Century 21’s website. In a lengthy judgment, Punnett J said:
“[127] Does the fact that Internet search engines access sites such as those of Century 21 using search engines, crawlers and robots change the legal situation? In my view it does not.
[128] The involvement of a machine in the contractual process was considered by Lord Denning long before the ubiquitous presence of personal computers and the Internet. ……
[129] A machine or a computer and the software that runs it has at some point been constructed and programmed by an individual. As noted by Sookman at 10.5:
… an electronic agent, such as a computer program or other automated means employed by a person, is a tool of that person. Ordinarily, the employer of a tool is responsible for the results obtained by the use of the tool since the tool has not independent volition of its own. When computers are involved, the requisite intention flows from the programming and use of the computer.
I agree with this statement. Liability is not avoided by automating the actions in question.”
For completeness, I note that the indexing by Zoocasa’s “vertical” search engine was not fully automated. The initial access was done manually by Zoocasa’s employee, but once the indexing was set up, it was automatically repeated daily.
123. In all the circumstances, the point of law is novel and of such importance that it should, in my opinion, be determined in its proper factual context at trial. I find it is an open question whether Google Search’s role in displaying Autocomplete and Related Searches suggestions upon entering a few keystrokes of an individual’s name is entirely passive such that it is a mere conduit or neutral distributor unaware of the presence of material that may have defamatory imputation, or whether they are subject to liability and responsibility as being more than just that.
124. The internet is constantly evolving, and it is to be expected there has been case law development since the hearing of the Summons. Whilst there is no need for me to rely on these further authorities in coming to my conclusion above, it may assist the parties in their future conduct of the present action if I refer to the recent Canadian case of Equustek Solutions Inc & anor v Morgan Jack & ors.[102]
125. In that case, the Supreme Court of British Columbia granted an interim injunction against Google Inc ordering it to stop indexing or referencing certain websites in its search results on a worldwide basis. In that case, the plaintiffs obtained court orders against the defendants that operated a “virtual company” and their websites were the subject of prohibitory court orders and yet they continued to sell their products on line. The plaintiffs sought the injunction against Google Inc (not a party to the legal action) on the basis that its search engine facilitated the defendants’ ongoing breach of the court’s orders.
126. Felon J in his judgment considered a number of issues in relation to internet search engines and had this to say about Google Inc’s argument that it merely offered a passive website for users who wished to search the internet:
“[47] Google submits that it merely offers a passive website to residents of British Columbia who wish to search the internet. It argues that its programs automatically generate search results without Google being actively involved in the particular search. Paragraph 23 of Google’s written submissions state:
[23] … Google’s internet search engine allows users to enter key-words and then Google generates a list of results in a specific ranked order. Google’s search results are computer generated through the use of Google’s highly confidential and proprietary algorithm and methodology. Google’s web crawler program (referred to as “Googlebot”) reviews the content that is available on trillions of webpages or URLs over the internet. Search results are generated based on that content [within seconds].
[48] I conclude that Google’s internet search websites are not passive information sites. As a user begins to type a few letters or a word of their query, Google anticipates the request and offers a menu of suggested potential search queries. Those offerings are based on that particular user’s previous searches as well as the phrases or keywords most commonly queried by all users. As James Grimmelman writes in “The Structure of Search Engine Law” (2007-2008) 93 Iowa L Rev 1 at 10-11:
Search engines are also increasingly learning from the large volumes of query data they have accumulated. A user’s history of queries can provide useful information about her probable intentions -- for example, whether she tends towards navigational or transactional queries. Similarly, search engines gain useful feedback into their own successes and failures by seeing which results users click on or by noticing long strings of searches on related terms, which may indicate that the user is having trouble finding what she’s looking for.
[49] Google collects a wide range of information as a user searches, including the user’s IP address, location, search terms, and whether the user acts on the search results offered by “clicking through” to the websites on the list.”
127. I further note that Google Inc has successfully applied for leave to appeal against Felon J’s judgment on the basis that the appeal cannot be said to be without merit and the issues on appeal are novel and important. However, Wilcock J of the Court of Appeal for British Columbia dismissed Google Inc’s application for interim stay as they were unable to demonstrate that irreparable harm would likely be incurred pending the hearing of the appeal.[103]
128. In Equustek, Google also raised arguments similar to Mr McCoy SC’s contentions:
“[136] First, Google says that it provides an important and valuable tool for navigating hundreds of trillions of webpages on the internet. Google argues it cannot, as a practical matter, monitor content or arbitrate disputes over content because of the enormous volume of content; because it cannot determine whether information is inaccurate or lawful; and because content on websites is constantly changing so even if Google could form judgments about the content of sites on its index at any given moment, those judgments would be obsolete moments later.”
But Felon J noted that the issue as to whether Google Inc is a passive indexer with no control over content has been the subject of litigation in other jurisdictions, including Trkulja v Google Inc and the judgment of the Court of Justice of the European Union in Google Spain SL and Google Inc v Agencia Espaňola de Protección de Datos (AEPD) and Mario Costeja González[104] (which grappled with a number issues including whether Google Inc was a “controller” of data obtained with the assistance of the Google search platform in Spain which resulted in answers that are mostly against the interests of Google Inc). [105] In Equustek, the arguments raised by Google Inc did not prevent Felon J from granting the interim injunction (although his decision is now pending appeal).
(c) Striking a balance
129. Mr McCoy SC drew my attention to article 27 of the Basic Law which provides inter alia that Hong Kong residents shall have freedom of speech, of the press and of publication, and also to article 19(2) of the International Convenant on Civil and Political Rights incorporated via article 39 of the Basic Law (see article 16 of the Hong Kong Bill of Rights Ordinance Cap 383 (“BORO”)). Article 16(2) of the BORO provides that “Everyone shall have the right to freedom of expression; this right shall include freedom to seek, receive and impart information and ideas of all kind ……”, but such right is expressly subject to article 16(3) which provides that “[the] exercise of the rights provided for in paragraph (2) of this article carries with it special duties and responsibilities. It may therefore be subject to certain restrictions, but these shall only be such as are provided by law and are necessary – (a) for respect of the rights or reputations of others; ……”
130. Mr McCoy SC submits that since the Human Rights Act 1998 the English courts have determined the test for continuing or adapting common law rules for the law of defamation is whether such rules would “impose a restriction … that amounts to a disproportionate restriction on freedom of expression”,[106] which approach has been upheld by the European Court of Human Rights in Times Newspapers Ltd v The United Kingdom (Nos 1 and 2).[107] In order to be proportionate, restrictions on the freedom of expression will need to be convincingly established by a compelling countervailing consideration, and the means employed would have to be necessary and proportionate to the ends sought to be achieved.
131. In Times Newspapers Ltd, the Times published two defamatory articles in its newspapers and on its website. The claimant sued for defamation and in the course of those proceedings; he brought a second action for defamation in relation to the continuing internet publication of the articles that remained on the defendant’s internet archive. The defendant complained of the chilling effect of the Duke of Brunswick rule that allowed multiple actions to be maintained for publication of the same article, which they claimed was a breach of their rights under article 10 of the Human Rights Convention. The European Court of Human Rights reiterated the importance of freedom of expression, but recognised that article 10 does not guarantee a wholly unrestricted freedom of expression even in matters of serious public concern, and the duties and responsibilities under article 10(2) assume significance when information imparted is likely to have a serious impact on the reputation and rights of private individuals.[108] In that case, the court found there was no violation of article 10, and the domestic court’s finding of libel was a justified and proportionate restriction to freedom of expression.
132. In Jameel (Mohammed) & anor v Wall Street Journal Europe Sprl,[109] Lord Bingham of Cornhill referred to article 10 of the Human Rights Convention which is reflected in section 12 of the Human Rights Act 1998, and said:
“18. Freedom to publish free of injustifiable restraint must indeed be recognised as a distinguishing feature of the sort of society which the Convention seeks to promote. ……
19. …… as the text of article 10 itself makes plain, the right guaranteed by the article is not unqualified. The right may be circumscribed by restrictions prescribed by law and necessary and proportionate if directed to certain ends, one of which is the protection of the reputation and rights of others. Thus a national libel law may, consistently with article 10, restrain the publication of defamatory material.
……
24. The tort of defamation exists to afford redress for unjustified injury to reputation. By a successful action the injured reputation is vindicated. The ordinary means of vindication is by the verdict of a judge or jury and an award of damages. ……”
133. In Crookes, the majority referred to the important role of the internet in promoting freedom of expression, and the importance of hyperlinks in facilitating access to information on the internet. Abella J said:[110]
“The Internet cannot, in short, provide access to information without hyperlinks. Limiting their usefulness by subjecting them to the traditional publication rule would have the effect of seriously restricting the flow of information and, as a result, freedom of expression. The potential “chill” in now the Internet functions could be devastating, since primary article authors would unlikely want to risk liability for linking to another article over whose changeable content they have no control. Given the core significance of the role of hyperlinking to the Internet, we risk impairing its whole functioning. Strict application of the publication rule in these circumstances would be like trying to fit a square archaic peg into the hexagonal hole of modernity.”
Quite apart from whether the move away from the strict publication rule is consistent with the reasoning in Fevaworks Solutions Ltd, Abella J also recognises the enormous damage that the internet can do to a person’s reputation by citing the following article by Professor Lyrissa Barnett Lidsky:[111]
“Although Internet communications may have the ephemeral qualities of gossip with regard to accuracy, they are communicated through a medium more pervasive than print, and for this reason they have tremendous power to harm reputation. Once a message enters cyberspace, millions of people worldwide can gain access to it. Even if the message is posted in a discussion forum frequented by only a handful of people, any one of them can republish the message by printing it or, as is more likely, by forwarding it instantly to a different discussion forum. And if the message is sufficiently provocative, it may be republished again and again. The extraordinary capacity of the Internet to replicate almost endlessly any defamatory message lends credence to the notion that ‘the truth rarely catches up with a lie’. The problem for libel law, then, is how to protect reputation without squelching the potential of the Internet as a medium of public discourse.”
134. Mr McCoy SC submits that it is essential to strike a proper balance between the right of freedom of expression and the need for protection of individual reputation, and the courts in Hong Kong should let the law of defamation develop in a manner which is compatible with constitutional rights. Mr McCoy SC says that since Google Inc cannot check or audit the truth or accuracy of the huge amount of information over which an automated search engine crawls, to hold Google Inc liable for their automatically and passively generated suggestions would be a disproportionate restriction on the freedom of expression bearing in mind (a) the extent to which relevant material has already become available to the public, (b) the effect of imposing liability on a search engine for signposting websites that publish defamatory material (ie it will make it more difficult for the public to find out about these websites), and (c) the important public interest in publication without being subjected to censorship consistent with the constitutional right to receive and impart information and which accords with the momentum and realities of the internet age and the needs of access to information and receipt of open knowledge.
135. The issue of the constitutional right of freedom of expression has been discussed in Fevaworks Solutions Ltd. Ribeiro PJ referred to the following observations by Lord Hobhouse in Reynolds v The Times Newspapers & ors and article 16 of the Bill of Rights to say that a purveyor of libellous statements cannot claim constitutional protection for his conduct:[112]
“…… There is no human right to disseminate information that is not true. No public interest is served by publishing or communicating misinformation. The working of a democratic society depends on the members of that society, being informed not misinformed. Misleading people and the purveying as facts statements which are not true is destructive of the democratic society and should form no part of such a society. There is no duty to publish what is not true: there is no interest in being misinformed. These are general propositions going far beyond the mere protection of reputations.”
136. But Ribeiro PJ also recognised the importance of society’s interest in freedom of speech and the free exchange of information and ideas (whether or not that information and those ideas find favour of any particular part of society), and the need to bear in mind competing social interests when approaching internet-related disputes:[113]
“111. …… The value of free and open many-to-many communications on discussion platforms must be recognized. The ability of internet intermediaries to host them in good faith must not be unduly impaired by the imposition of unrealistic or overly strict standards which would make commercial operation impossible or introduce a chilling effect discouraging free and open exchanges. At the same time, a platform provider must genuinely recognize and take all reasonable steps to protect the rights and reputations of persons from being unlawfully damaged by postings published on the forum. ……”
137. Mr Ng submits the Words are extremely serious and highly defamatory of Yeung.[114] Google Inc concedes that for the purpose of the Summons they will not take issue that the Words are defamatory of Yeung. As a matter of law, a defamatory imputation is presumed to be false,[115] so Mr Ng says that for the purpose of the Summons the Words are presumed to be false. Mr Ng also reminds that by virtue of the 1st Letter, Google Inc has knowledge of the presence of the Words on the Websites, and it is proportionate to require them to take down the offending Words after notice, which achieves the proper balance between the freedom of expression and the protection of reputation.
138. Mr McCoy SC submits that since Australia, unlike Hong Kong and Europe, has no Bill of Rights and its constitution does not deal with the substantive rights relating to freedom of expression and freedom to receive and impart information, the local position is distinguished from Trkulja v Google Inc which is a decision by a single judge in Australia, and I should read that judgment with caution.
139. However, I note that in A v Google New Zealand Ltd, the defendant invoked section 14 of the New Zealand Bill of Rights Act 1990 and argued that the development of the law in this area ought to be consistent with the right of freedom of expression. It was said that “it would be an unreasonable limit upon that right to hold that the automatic generation of search results amounts to publication. Search engine operators are, according to the defendant, an invaluable gateway to the internet, performing a vital role in the dissemination of information. Liability as a publisher would have a “chilling” effect and would thus adversely affect users’ access to information”. Abbott J said:
“[72] Whilst the right to freedom of expression as protected by the NZBORA ought to be considered in the development of the law in this area, it may not be an unreasonable limit upon that right to hold that a search engine is a publisher of both specific URLs and words that appear in snippets (which search engine providers have chosen to include and which elevate hyperlinks beyond the status of mere footnotes).
[73] This could mean that search engine providers would be responsible where an offending hyperlink is deactivated but its snippet continues to appear, thus addressing one of the plaintiff’s concerns. Such an approach is consistent with the broad common law definition of “publication” as being the communication of a statement to just one other person. To limit this definition to exclude the repetition of information where that repetition occurred without human input unnecessarily confuse this area of the law. It may therefore be more appropriate to hold that a search engine is a publisher but with access to the defence of innocent dissemination. The possibility of a defence in innocent dissemination where the defendant has not had notice of the defamatory material may be sufficient recognition of the fact that a search engine is a neutral index. However, these decisions need to be made with all available facts before the Court and with the benefit of legal argument more specifically focussed on the points than was possible on this application.
[74] The plaintiff’s concern that removal of specific web pages and deactivation of the hyperlink appears to be an impotent response may turn out to be a matter more suited to determination by legislation.”
140. I share Abbott J’s observations, and find that in the search for the proper balance between freedom of expression and protection of reputation, there is debate as to whether holding Google Inc to be a publisher of the Autocomplete and Related Searches suggestions under the broad traditional concept of “a publisher” is an undue restriction of the right to freedom of expression, especially when it is open for Google Inc (if they can) to raise the defence of innocent dissemination on the basis, as they claim, they have no awareness of any defamatory material. I also agree that any decision on this should be made in the proper factual context and with the benefit of legal argument on the elicited facts. In the circumstances, these considerations do not dissuade me from my view that there is a good arguable case to Yeung’s claim.
(d) Summary
141. To conclude, there is a good arguable case that Google Inc is the publisher of the Words and liable for their publication. With Hong Kong being the place of publication, it follows therefore there is also a good arguable case that damage has been sustained within this jurisdiction or the damage has resulted from an act committed within the jurisdiction. It also follows that Hong Kong is the appropriate forum.
VIII. SERIOUS QUESTION TO BE TRIED
(a) Real and substantial tort or Jameel abuse of process
142. On whether there is a serious question to be tried, one argument advanced by Mr McCoy SC is that there is no evidence of real and substantial readership in Hong Kong capable of damaging Yeung’s reputation. He says the evidence that Yeung’s staff and/or legal team have accessed the Words on the Websites merely demonstrates that such access is possible but shows there is very limited search activity in Hong Kong and none to a genuine third party. Clark says that unless there is evidence of substantial readership in Hong Kong, there is no or no significant damage to Yeung’s reputation. In such circumstances, the potential damage (if any) is too trivial to merit maintenance of these proceedings in Hong Kong. Mr McCoy SC cites Jameel (Yousef) to say that the Leave Order should be set aside on the basis that “the game is not worth the candle”.[116]
143. Mr Ng says there is no basis for Google Inc to say there is no substantial damage to Yeung’s reputation because in a libel action general damage is presumed.[117] But as explained in Gatley, the presumption of damage must now be qualified in two respects, one statutory which is not relevant to this jurisdiction and the other is Jameel abuse of process.[118] In Jameel (Yousef), the Court of Appeal decided that if a claimant brought an action in circumstances where his reputation had suffered no, or minimal, actual damage, it may constitute an interference with freedom of expression that is not necessary for the protection of the claimant’s reputation, which may render the action susceptible to being struck out as an abuse of process.
144. But Mr Ng goes further to say that Jameel abuse of process has no relevance to Hong Kong because the Hong Kong courts have declined to follow the English approach. He refers to Wong Wing Ho v Chong Lai Wah & anor,[119] but in that case Tang VP (as he then was) rejected such approach on the basis that at that time Hong Kong had not adopted rules similar to the Civil Procedure Rules and the implication of the English approach had not been canvassed.[120] However, since Wong Wing Ho, Hong Kong has undergone the Civil Justice Reform. In Jotron AS (formerly known as Jotron Electronics AS) v Stanley Chang trading as Hong Kong Instech Trading Co,[121] Sakhrani J distinguished Jameel (Yousef) on the basis that the court there was dealing with inter alia the summary disposal of the claim pursuant to section 8(2) of the Defamation Act 1996, which provides that the court may dismiss the claim if it appears to the court it has no realistic prospect of success and there is no reason why it should be tried. There is no similar provision in Hong Kong.
145. It is true that Jameel abuse of process was brought up in a striking out application by the defendant in Jameel (Yousef). This is not directly relevant here because the Summons does not seek to strike out Yeung’s claim against Google Inc. However, Mr McCoy SC’s complaint of minimal publication and no substantial damage remains relevant in the context of an application to set aside service outside jurisdiction on the basis that the alleged libel does not reach the threshold of seriousness required to establish a real and substantial tort that has been committed within the jurisdiction. This was the alternative approach in Jameel (Yousef).[122] Indeed, there is jurisprudence on this approach since as early as in 1937: Kroch v Russell et Cie Société des Personnes à Responsibilité Ltd[123] and Chadha v Dow Jones & Co Inc[124] cited in Jameel (Yousef).[125] Indeed, in the pre-Jameel local case of Investasia Limited & anor v Kodansha Co Ltd,[126] which has not been cited by counsel but which Clark must be familiar since he was also professionally involved in that case, a Japanese national was permitted to sue two Japan-based defendants in relation to material published in Japanese both on the internet and in a magazine with very limited distribution in Hong Kong. The learned judge considered the plaintiff had sufficient business connections with Hong Kong (including spending 4-6 months a year here) to have established a reputation here that deserved protection. “If a plaintiff has a reputation in Hong Kong, as the plaintiffs in this case have undoubtedly established they have, it is not right to tell him to go elsewhere to vindicate that reputation.”[127] In that case, 500,000 copies of the magazine were published of which 157 copies were distributed in Hong Kong. Findlay J regarded the 157 copies distributed in Hong Kong as a significant number, especially when the alleged defamatory material was sensational and juicy and in the nature of things likely to spread from mouth to ear quickly amongst those who might do business with the plaintiff, and said:
“14. …… Of course, one can do sums with the figures and say that the number published in Hong Kong compared with the total number is very small, but I do not think this is the right approach. What I am concerned about is whether or not the number is significant for the purposes of deciding whether a complaint about the publication could be said to be one of substance. I think it is.”
146. Since Google Inc in the Summons does not ask for Yeung’s claim to be struck out, whether for abuse of process or otherwise, I am prepared to consider Mr McCoy SC’s core complaint of minimal publication and no substantial damage as going to the question of whether there is any real and substantial tort that justifies service out of jurisdiction, bearing in mind the underlying objectives of the Civil Justice Reform and the right to freedom of expression.
147. Mr Ng’s contention of a real and substantial tort premised on substantial publication/damage rests on three matters: (a) there is sufficient basis in the Hom 1st Aff for the court to infer substantial publication, (b) the Yuen Aff provides ample evidence of substantial publication, and (c) in view of the seriousness of the defamatory imputations of the Words, character vindication and not money is a significant objective.
148. I start with a short description of the statistical calculations in the Yuen Aff. Yuen explains that by using Google AdWords[128] and Google Trends[129] a registered user can obtain the monthly average number of “google” searches done for any particular keyword(s) in any specified location (ie the Monthly Search Volume) and the popularity of any particular keyword relative to the total number of “google” searches over time (ie “Relative Popularity” or index score), and then calculate the estimated number of “google” searches conducted against any particular keyword(s) in a specified location for any selected month (within the period of the past 12 months).
149. On 6 February 2013, Yuen accessed Google AdWords and Google Trends to obtain the Monthly Search Volume and Relative Popularity of the search Keywords, and made calculations as to the estimated number of “google” searches for such keywords in Hong Kong for July 2012. The “Local Monthly Searches” (每地當月搜尋量)[130] for “楊受成”, “Albert Yeung” and “Albert Yeung Sau-Shing” in Hong Kong are 6,600, 1,000 and 260 respectively. By multiplying such “Local Monthly Searches” by 12, the estimated annual search volume of such Keywords are 79,200, 12,000 and 3,120 respectively. The Relative Popularity or index score for the keyword “楊受成” for July 2012 and for the last 12 months from July 2012 are 45 and 448 respectively.
150. Based on such information, the estimated number of “google” searches made for “楊受成” in Hong Kong for July 2012 (being the full month immediately before the WoS was issued) is calculated as 7,955 pursuant to a set formula. But the Relative Popularity of keywords “Albert Yeung” and “Albert Yeung Sau-Shing” for July 2012 are not available from Google Trends because there is “not enough search volume to show graph”, so Yuen cannot calculate the estimated number of “google” searches for such keywords for July 2012. But the approximate 12-month average number of “google” searches for such keywords can still be gathered from Google AdWords. The Local Monthly Searches for keywords “Albert Yeung” and “Albert Yeung Sau-Shing” are 1,000 and 260 respectively.
151. For the purpose of the Summons, Google Inc does not take issue with the numbers produced by Yuen as being applicable to searches performed on the .com.hk Website, but Mr McCoy SC raises the following contentions:
a) The Yuen Aff that seeks to analyse/calculate the estimated number of “google” searches made in Hong Kong for the Keywords has not been placed before the court at the ex parte stage, and is only presented in answer to the Summons to patch up the inadequacy of the Hom 1st Aff at the ex parte stage.
b) Even on Yeung’s best case, there were only about 8,000 “google” searches for “楊受成” in July 2012. The 7,995, 1,000 and 260 searches against the Keywords are “trivial” when compared with more than one billion searches conducted daily using Google Search and also with the daily circulation of 251,895 copies and 108,047 copies for Apple Daily and South China Morning Post respectively.
c) The same person can conduct the same search multiple times thus skewing any count of search results, so the fact that 7,995 searches have been done does not mean that 7,955 separate users performed and saw the results of 7,955 searches.
d) Just because users searched the Keywords does not mean they saw the predictive suggestions complained of. It cannot be said with any degree of certainty that on a given day the Autocomplete and Related Searches suggested results for particular search terms will be the same as those conducted on another occasion, and Yeung has not provided evidence of the number of times the predictions have been displayed.
On such basis, Mr McCoy SC says there cannot be any inference of real or substantial publication.
152. Mr Ng says the evidence as to searches done by the IT Dept and ILL and/or the statistical evidence in the Yuen Aff demonstrate that (a) when searches were made with the Keywords at the Websites, Google Search would instantaneously and automatically generate the Words as autocompletions and related search terms, and (b) there is inference for substantial publication of the Words which were available for downloading on 6 August 2012 by all members of the public in Hong Kong who had access to the internet. As explained in the Hom 2nd Aff, Yeung has significant connection with, and a reputation to protect, in Hong Kong. Since “googling” is a commonplace mode to gather information nowadays, it is highly probable that any person who wants to do business with or find out more about Yeung would “google” him, which service is free to anyone who has access to the internet, and the Words would have been accessible to anyone who typed Yeung’s name into the Search Box on the Websites.
153. Mr Ng submits that on this basis together with Yuen’s statistics that 7,995, 1,000 and 260 searches had been made against the Keywords in July 2012 alone, the inference of substantial publication to third parties is irresistible. He adds that since Google AdWords and Google Trends are Google Inc’s own internet service tools, Google Inc is well-positioned to acquaint themselves with information they possess of the number of searches made against the Keywords over any given period. Instead, they have chosen to withhold such information and to file the Summons, which is against the spirit of the underlying objectives in Order 1A of the RHC to ensure fairness between the parties.
(b) .com and .com.tw Websites
154. Before I deal with the scale of publication, I shall address a matter raised in the Clark 1st Aff. Clark claims that notwithstanding the statistics compiled by Yuen, Yeung has not produced any evidence that anyone in Hong Kong had accessed the .com.tw or the .com Websites and performed searches using the Keywords. Mr McCoy SC says that even on Yeung’s own case there appears to be no actionable claim capable of being maintained by Yeung in relation to the Words on the .com.tw and .com Websites.
155. The Clark 1st Aff explains that the rationale for the different domain names and the Websites linked to them is that they are country specific. Users who navigate to the .com Website from an IP address Google Inc believes is in a different country will be redirected to the appropriate domain for the country they are located in, eg a user from Hong Kong who seeks to navigate to the .com Website will be redirected to the .com.hk Website, and a user in Taiwan who seeks to navigate to the .com Website will be redirected to the .com.tw website. This redirection can be manually overridden by the user by specific action to prevent such redirection from happening again. This country specific service and redirection process is not applied to every country worldwide; in some countries where internet service is poor this service/process is not applied. The purpose of such redirection is to make search results more relevant to users who will get a general product experience appropriate to their country. The search quality algorithms consider a user’s country (as specified by domain) and work out what content is likely to be most relevant to that user. A user of, say, the .com.tw website will see content that is relevant to his/her location, eg if a search is conducted for pizza from a IP address Google thinks is in Taipei, map results with pizza restaurants in Taipei will be shown. The same is not true if the same user, located in the same place, uses the .com Website.
156. The Clark 1st Aff further explains that a “member of the public in Hong Kong on 6.8.2012 who had access to the internet” would not be provided with search predictions or search results from the .com and .com.tw Websites without deliberately altering his/her settings so as to prevent redirection to the com.hk Website. He complains that the SoC and the Hon 1st Aff do not distinguish between the Words provided as search predictions and/or related search items by reference to each of the three Websites. Therefore, it cannot be inferred, particularly with regard to the .com and .com.tw Websites, that there was actual publication to a third party in Hong Kong, ie an act committed in this jurisdiction or damage sustained in this jurisdiction, without actual evidence from a third party who has deliberately bypassed Google Inc’s redirection function and accessed these two Websites.
157. Mr McCoy SC submits that given Clark’s explanation of automatic redirection of Hong Kong users to the .com.hk Website and the absence of evidence that someone in Hong Kong has actually re-directed himself from the .com.hk Website to the .com and .com.tw Websites, none can be inferred because manual intervention is required. In the circumstances, there is no serious question to be tried in respect of Yeung’s claim concerning the Words on the .com and .com.tw Websites.
158. I agree with Mr Ng’s three-fold answer. First, the Hom 1st Aff and Yuen Aff confirm that searches have been conducted in Hong Kong against all three Websites and not just the .com.hk Website.
159. Secondly, Clark himself confirms that such redirection can be manually overridden so that the automatic redirection will not happen again, which confirms that the .com.hk and .com Websites can be assessed from Hong Kong. Indeed, since Google Search can be used by entering the name of any of the worldwide domains into a web browser, there should be no logical difficulty for an internet user simply to type a search query in the Search Box to look for the .com Website and/or .com.tw Website, and upon access to carry out internet search at such Website(s) without having to alter settings to prevent all future automatic redirection to the .com.hk Website.
160. Thirdly and more significantly, Mr Ng argues that such explanation by Clark is inadmissible because he has failed to expressly stipulate the sources and grounds of his information and belief in respect of his assertion that search queries made in Hong Kong are automatically redirected to the .com.hk Website as required under Order 41 rule 5(2) of the RHC. Mr Ng reminds that Clark is Google Inc’s solicitor, who does not say he has personal knowledge of the technical operation of the Google Search engine, which matters should logically be within the knowledge of Google Inc.
161. It has been said that Order 41 rule 5(2) of the RHC should be strictly complied with where objection is likely to be taken,[131] and that an affidavit of information and belief, not stating the sources of the information or belief, is irregular and therefore inadmissible as evidence, whether on an interlocutory or final application.[132] In Korea Sonbak Shipping Co v Charter Harvest Shipping Ltd,[133] whilst affirming Leung Kin Fook & ors v Eastern Worldwide Co Ltd,[134] the court took a pragmatic approach where a deponent of an affidavit used in interlocutory proceedings deposed that certain documents supplied to him were the source of his information and belief. It is said that Order 41 rule 5(2) of the RHC has been complied with even if they are not exhibited to his affidavit provided that, if he is required to produce the documents, he either does so or provides an explanation for not doing so which the court regards as acceptable.[135]
162. In my view, there is force in such argument. When Google Inc issued the Summons, given the nature of the application made under Order 12 rule 8 of the RHC which if successful puts an end to Yeung’s claim in this jurisdiction, Google Inc would have expected it to be contested. Since the subject matter of automatic redirection of search queries by Google Inc is raised by them and is not a matter that arises from Yeung’s case, it is incumbent on Google Inc to abide by the rules in order to properly place before the court the explanation they now seek to put forward. It is therefore necessary for Clark to comply with Order 41 rule 5(2) of the RHC, indicating with reasonable particularity the sources of his information. It cannot be said that Google Inc is unaware of such duty for (a) the Clark 1st Aff exhibits documents to support Clark’s explanation as to the operation of Google Inc’s search engine, and (b) the Clark 2nd Aff specifies that where facts and matters in that affidavit are not within Clark’s personal knowledge his source of knowledge is Henry Lien, associate product counsel employed by Google Inc. Google Inc has therefore failed to adduce admissible evidence to suggest there is no search activity with respect of searches made with the Keywords for the .com and .com.tw Websites.
(c) Scale of publication
163. This nicely brings me to the question of whether the scale of publication of the Words in Hong Kong is so minimal and the reputational damage so trivial that there is no real and substantial tort to support the requisite seriousness to justify service out of jurisdiction.
164. Publication is mixed question of fact and law, and it must depend on the circumstances of each case whether or not publication has taken place.[136] In internet cases, there is no presumption of substantial publication.[137] But where the claimant proves that the impugned material has been published to one or more persons, it may be inferred that the material (if easily accessible) has been published more widely within the jurisdiction.[138]
165. In Loutchansky v Times Newspapers Ltd (No 2),[139] the claimant sued for libel in respect of two articles published in The Times and its website which accused him of being engaged in serious international criminal activities. Gray J approved the following passage from Gatley: “It is not necessary in all cases to prove that the libellous matter was actually seen and read by some third party. If it is a matter of reasonable inference that this has happened a prima facie case of publication will be established.” [140]
166. In Steinberg v Pritchard Englefield & anor,[141] the claimants sued over a letter published by the defendant on the internet that alleged they artificially and unprofessionally inflated their solicitor and client costs. That letter was accessible to anyone who fed the claimants’ names into a standard search engine and readable by anyone who accessed the defendant’s own professional website. Sedley LJ held that in those circumstances, the inference of substantial publication was irresistible.
167. In Gregg v O’Gara,[142] a libel claim was brought in respect of two internet articles published by the defendant from his home in Ireland. The claimant applied for summary judgment, and on the basis of uncontroverted evidence that the articles were accessible by typing “Yorkshire ripper” into a standard internet search engine and that the Yorkshire ripper case was a topic of continuing public interest, King J held any jury would draw the irresistible inference that the two articles had been widely published within the jurisdiction.
168. In Mardas v New York Times Company & anor,[143] a case not cited by either counsel, the claimant brought proceedings against the New York Times and the International Herald Tribune in an English Court for publishing an allegedly defamatory article. Evidence established there were approximately 177 hard copies and 31 hits of the story on line. Eady J refused the defendant’s suggestion of insufficient publication within the jurisdiction by relying in the remarks of Callinan J in Gutnick that “[a] publisher, particularly one carrying on the business of publishing, does not act to put the matter on the internet in order to reach a small target. It is its ubiquity which is one of the main attractions to users of it. And any person who gains access to the Internet does so by taking an initiative to gain access to it in a manner analogous to the purchase or other acquisition of a newspaper, in order to read it ……” Eady J noted that the article complained of was published online and continued to be available online, which fact “naturally gives rise at least to the possible inference that there has been a continuing, albeit modest, readership”. He allowed the action to proceed. In discussing whether there has been a real and substantial tort within the jurisdiction (or, at that stage of the proceedings, arguably so), Eady J said “[this] cannot depend on a numbers game with the court fixing an arbitrary minimum according to the facts of the case”.[144]
169. In Trkulja v Yahoo! Inc LLC, the defendants conceded that the defamatory article had been downloaded and read by at least one person using the Yahoo! 7 internet search service and they had published the article such person(s), but they argued that the article would not have been read by the world at large and instead would have been read by people who specifically wanted to know about the plaintiff. Kaye J observed that “…… he would be entitled to take into account, as a matter of judicial notice, that the use of the internet, to ascertain information about particular people, is now commonplace, Indeed, those searches have entered the everyday lexicon of the community, and the process of undertaking such a search, is commonly referred to as “to google”. Further, …… the people, to whom the defendants published the article, would have used the defendants’ search engine specifically to obtain information about the plaintiff. ……”[145] Kaye J noted that the evidence established that the material remained available to be accessed on the internet through the defendants’ search engine until the end of 2010, which could only serve to increase the damage to the plaintiff’s reputation throughout the community. It was said that the publication had a more injurious effect than if it had been made on the live or print media since it was open to persons, who heard about the publication, to view it for themselves and to confirm the contents/effect of it. On such basis, Kaye J was satisfied the publication of the article caused significant grief and distress to the plaintiff.
170. Further, “[the] fact that a snippet from and link to an online statement appears on the first screen of results when a search is conducted on the claimant’s name may be sufficient to support an inference that a substantial number of persons have searched for and read the information appearing the snippet”.[146]
171. Turning to the publication of the Words, there is affirmation evidence before me that in both May and August 2012 “google” searches at the Websites using the Keywords generated the very same 1st, 2nd, 3rd and 4th Words as autocompletions and/or related search terms. Since such results are driven by previous search activity and/or web content, the matching results in May and August 2012 lend themselves to an inference of similar search/web interest during the Period. It is Google Inc’s own case that the popularity and volume of previous search activity using the relevant keywords that drive the autocompletions generated by Google Search, and there has been no other explanation for why the 1st and 2nd Words appear so instantaneously in the drop-down menu of the Search Box in both May and August 2012. The same can also be said for Related Searches results although they are also influenced by web content.
172. Taking into account the contextual background of Yeung’s involvement in the more flamboyant businesses of entertainment and gaming, the subject matter of the Words being topics of continuing interest to members of the public in this jurisdiction in view of the nature of autocompletions and related searches as explained above, Google Search being a popular search engine such that it is now commonplace to “google” about persons, the ubiquity of the internet and the ready accessibility of the Autocomplete and Related Searches suggestions to anyone conducting searches by Google Search, it appears to me to be an open question of fact, by inference or otherwise, that visits have been paid to the Websites and as a result of such visits being paid there was access to the Words. It seems to me to be the kind of matter that should be determined at trial rather than on an application to set aside service out of jurisdiction. The fact that the same material remained available as autocompletions and related search terms over the Period without being replaced by other words arguably serves to fuel the damage to Yeung’s reputation in Hong Kong, which may well have more injurious effect than if the Words were available from live or print media.
173. In my view, even on the basis of the Hom 1st Aff, it is an open question by inference whether there is real and substantial publication of the Words. In the circumstances, there is no material non-disclosure at the ex parte stage. The principles in respect of duty of disclosure have been set out in Hong Kong Civil Procedure 2014,[147] and I have summarised them in China Agri-Products Exchange Limited v Wang Xiu Qun.[148] I do not propose to repeat those principles save to say that non-disclosure is culpable only when the information goes to the weighing operation of the ex parte judge or master, ie it must be something which would have affected his decision on the application.
174. Even if there is material non-disclosure sufficient to discharge the Leave Order (which I disagree), I must go on to “consider the matters not disclosed on the original application and, if appropriate, re-grant leave on terms that the plaintiff pay all the costs of the discharge application”, and such decision whether or not to renew leave is obviously discretionary.[149] But where there is no intention to deceive and any re-application would be bound to succeed (ie the result with disclosure would have been the same), the court will not discharge the leave. “Where material non-disclosure has been established and the case is otherwise a proper one for service out, the court will now need to apply the principle of proportionality in weighing the gravity of the omission against the proposed impact on the proceedings in terms of discharge or costs.”[150]
175. In the circumstances, I turn to the statistics obtained by Yuen in case I am wrong and there is material non-disclosure. In my view, the inference of publication as discussed above is bolstered by the statistics compiled by Yuen. Mr McCoy SC raises the spectre of the same person doing multiple searches of the Keywords and the possibility of variability in the Autocomplete and Related Searches response. But these assertions are grounded on possibility and not fact. This seems to be the kind of argument that could be developed at trial rather than on the present application.
176. Mr McCoy SC also refers to the triviality of the 9,000 or so searches by the Keywords in July 2012 alone compared with the massive volume of worldwide internet searches using Google Search or with the local print media. I agree with Eady J that this is not a numbers game. The present action is brought by a businessman well-known within this jurisdiction, and Mr McCoy SC in his submissions has not challenged Yeung’s claim that he is a man with a reputation to protect in Hong Kong. In my view, it cannot be said at this stage that damages for reputational damage in Hong Kong are likely to be minimal if Yeung wins at trial. The allegations in the Words have criminal associations (and for the present purpose Google Inc does not take issue that the Words are defamatory of Yeung), and an award may take into account the matters Yeung relies in aggravation.[151] A jury (or judge if the matter is to be tried by judge alone) may consider that such returns by the Autocomplete and Relates Searches features made publicly available to all internet users and persisted after being served with take-down notification merit a substantial award. It is trite that an award of damages in libel actions serves not only to vindicate the reputation of the claimant, but also to compensate him for injury to his feelings.
177. Looking at the matter in the round, even if there had been non-disclosure (which I do not agree), I would have either not set aside the Leave Order or re-granted the Leave Order. I am unable to conclude on the affidavit evidence before me that Yeung will have nothing to gain from this action or that the potential benefits to him of continuing with this claim are so outweighed by the countervailing considerations that I should set aside service out of the jurisdiction. In all the circumstances, I am persuaded there is a serious question to be tried.
(d) Alternative claims
178. The Clark 2nd Aff states that the whole purpose of an internet search query is to look for content relating to the search terms. Predictive suggestions are tools for enhancing Google Search’s functionality and do not provide content of the relevant websites which are the true answers to the search queries. Google Inc cannot intervene and change the URLs of such third party websites which content has been uploaded to the internet. Google Inc neither creates nor hosts the relevant webpages, but the third parties responsible for the relevant websites can remove content to other webpages and/or render their URLs of no use so as to block access to such content. In the circumstances, it is said that Yeung does not need a remedy against Google Inc in order to obtain access to justice because he has an effective claim/remedy against the originators and operators of the individual websites from which the autocompletions and/or related search terms are derived.
179. Mr McCoy SC suggests that Yeung should not adopt the complicated and unprecedented action to claim against Google Inc premised on an artificial responsibility as alleged publisher of the Words on the Websites, and ignore the easy and obvious cause of action against the relevant websites and if necessary to make Norwich Pharmacal discovery applications to identify the parties responsible for or who hosts or who places information on those websites. Mr McCoy SC posed the Armageddon scenario with everyone unrelentingly suing every search engine for every uncomplimentary remark or result extracted from previous search query or any indexed webpage. He reminds that even if Google Inc were to remove the suggestions, it does not actually remove the underlying information from the internet.
180. On the other hand, Mr Ng submits (and I agree) this is a red-herring argument since the Words are published by Google Inc on the Websites owned/operated by them and not by third parties on third party internet platforms. The Words are the results of recombinations, aggregations and synthesis of previous search activity and web content by Google Search’s algorithms as designed by Google Inc. Google Inc does not merely convey information from computer to computer, and it is at this stage unclear how the suggestions actually differ from web content or previous search terms, and it is not easy at this stage to appreciate how Yeung can pursue specific website owners/operators or individual internet users for reconstituted and aggregated suggestions. Further, as discussed above, Google Inc does have some ability to censor material. Still further, even though it may be argued that it is ultimately the user’s choice in selecting or keying in his search query in the Search Box, autocompletions and related search terms do not just point to a likely search query, but by its immediate and ready access invites the user towards them. It cannot be said at this stage there is no point to Yeung’s claim against Google Inc irrespective whether or not he has other claims against specific websites.
181. Mr McCoy SC says the availability of a cause of action against specific websites was regarded as significant in Bunt:[152]
“79. The claimant is not deprived of access to justice. His remedies lie against the first to third defendants (if he can establish the necessary ingredients in respect of each). They may not be persons of substance, such as to make it worthwhile pursuing them. Even if that is right, it is clearly not a sufficient reason for bringing in the present applicants. I therefore accede to the submissions of the applicants and will grant orders accordingly.”
I do not think this supports Mr McCoy SC’s contentions. Quite apart from the fact that Bunt has to be read in light of Fevaworks Solutions Ltd as explained above, Eady J has himself explained his observations in Tamiz v Google Inc & anor as follows:[153]
“…… In Bunt v Tilley, cited above, the claims against AOL UK Ltd, Tiscali UK Ltd and British Telecommunications Plc were all struck out. Against that background, I concluded the judgment with these words:
‘The Claimant is not deprived of access to justice. His remedies lie against the First to Third Defendants (if he can establish the necessary ingredients in respect of each). They may not be persons of substance, such as to make it worthwhile pursuing them. Even if that is right, it is clearly not a sufficient reason for bringing in the present applicants. …’
The important question, in the present context, would appear to be, not whether Mr Tamiz can identify the authors or bloggers in question, still less whether they are worth powder and shot, but rather whether he is in a position to establish against Google Inc the necessary attributes of a publisher in accordance with common law principles.”
This I agree, and I have found there is a good arguable case that Google Inc is a publisher in respect of the Words on the Websites as alleged by Yeung.
(e) Previous litigation
182. Mr Ng places much emphasis on the previous litigation between Yeung and Google Inc in HCA483/2011, which I have tried to summarise in footnote 3 above. I agree with Mr McCoy SC that it is immaterial to the issues herein, and do not propose to discuss this any further.
(f) Summary
183. In all the circumstances, I consider there is a serious question to be tried, and Master K Lo was justified in exercising her discretion to grant the Leave Order.
IX. CONCLUSION
184. I therefore dismiss the Summons, and grant a costs order nisi in favour of Yeung against Google Inc (with all costs reserved if any) to be taxed if not agreed.
185. I return to the beginning of this Decision and say that the internet has become a universal medium. The advantages of having easy access to a rich store of information are many, and they have been widely applauded. But such benefit comes at a price; any risk of misinformation can spread easily as users forage in the web. The art is to find the comfortable equilibrium in between.
186. I take this opportunity to thank counsel for their assistance.
(Marlene Ng)
Deputy High Court Judge
Mr Lawrence Ng, instructed by Iu, Lai & Li, for the plaintiff
Mr Gerard McCoy SC, instructed by Deacons, for the defendant
-----------------------
[1] translated as “Yeung Sau Shing Triad”, “Yeung Sau Shing Perverted” and “Yeung Sau Shing Sun Yee On”
[2] translated as “Yeung Sau Shing Triad”, “Yeung Sau Shing 10 Sins”, “Yeung Sau Shing Sun Yee On”, “Yeung Sau Shing 14k”, “Yeung Sau Shing covered story” and “Joey Yung Yeung Sau Shing covered story”
[3] on 23 May 2011 Yeung commenced HCA483/2011 against Google Inc for damages for libel in respect of defamatory material on the Blogspot website operated, managed, administered and/or hosted by Google Inc, but on 25 August 2011 Google Inc by their solicitors advised that the blog was no longer publicly available, and by consent Yeung agreed to withdraw the action with no order as to costs
[4] legal counsel to the Group
[5] associate director of the Group’s corporate communication department
[6] the Hom 1st Aff draws attention to Gatley on Libel and Slander 11th ed para 6.1 at pp 187-189 and Emperor (China Concept) Investments Ltd v SBI E-2 Capital Securities Limited & anor HCA2686/2004 (unreported, 10 January 2006) para 16 (see also Oriental Press Group Ltd & anor v Fevaworks Solutions Lted [2013] 5 HKC 253, 277 at para 68)
[7] see paragraph 15 above
[8] see Dow Jones & Co Inc v Gutnick (2002) 194 ALR 433 cited in Emperor (China Concept) Investments Ltd para 16 and Godfrey v Demon Internet Ltd [2001] QB 201, 208-209
[9] (1849) 14 QB 185, 188-189
[10] [1981] 1 QB 524, 527
[11] [2001] QB 201, 208
[12] see also Collins on Defamation (2014) para 4.04 at p 70 and Gatley on Libel and Slander 12th ed (2013) para 6.15 at pp 207-208
[13] [2005] QB 946
[14] para 6.01 at p 189
[15] HCA3831/1997, Chu J (as she then was) (unreported, 14 November 2000) p 9
[16] HCA3831/997, Chu J (as she then was) (unreported, 15 November 2000) p 3
[17] CACV1006/2000 (unreported, 17 April 2002) para 34
[18] see Gatley para 3.20 at pp 129-130
[19] see Gatley para 3.22 at pp 132-133
[20] [1937] 1 KB 818 (see paragraphs 60-62 below)
[21] [2013] EWHC 2694 (5 September 2013) para 43 cited in Gatley para 6.29 (footnote 250) at p 227
[22] see Trkulja v Google Inc LLC & anor [2012] VSC 533 (12 November 2012) para 18
[23] [2013] 5 HKC 253, 262-263 at paras 18-19
[24] p 264 at para 23
[25] Fevaworks Solutions Ltd pp 266 and 283 at paras 29 and 90
[26] p 273 at para 56
[27] [1937] 1 KB 818
[28] Fevaworks Solutions Ltd p 271 at paras 46-48
[29] at pp 837-838
[30] at pp 834-835
[31] [1991] Aust Torts Rep 81-127 (22 December 1988)
[32] pp 270-271 at para 44
[33] Fevaworks Solutions Ltd pp 271-272 at para 49
[34] Fevaworks Solutions Ltd pp 272-273 at paras 50-54
[35] [2013] EMLR 14 at p 308
[36] [2013] NZHC 540 (25 October 2012)
[37] Fevaworks Solutions Ltd pp 272-273 at para 53
[38] Fevaworks Solutions Ltd pp 272-273 at para 53
[39] Chau Hoi Shuen, Solina Holly v SEEC Media Group Limited CACV275/2012 (unreported, 25 April 2014) para 5.3
[40] Fevaworks Solutions Ltd pp 275-276 at para 63
[41] Fevaworks Solutions Ltd p 279 at para 76
[42] Fevaworks Solutions Ltd p 283 at para 89
[43] see paragraph 65(b) above
[44] Fevaworks Solutions Ltd pp 278-279 at paras 74-76
[45] Fevaworks Solutions Ltd p 263 at paras 19-20
[46] see Gatley para 6.26 at pp 217-218
[47] see Gatley para 6.26 at pp 217-218 and Collins at para 4.60 at p 84
[48] [2002] QB 783
[49] see Collins para 4.62 at p 85
[50] [2007] 1 WLR 1243 and Fevaworks Solutions Ltd p 280 at para 82
[51] at p 1249
[52] at p 1252
[53] see Oriental Press Group Ltd v Fevaworks Solutions Ltd [2012] 1 HKLRD 850, 870
[54] Fevaworks Solutions Ltd p 280 at para 81
[55] Fevaworks Solutions Ltd p 281 at para 84
[56] Fevaworks Solutions Ltd p 281 at para 84
[57] Fevaworks Solutions Ltd p 280 at para 81
[58] para 6.27 at pp 221-222
[59] Fevaworks Solutions Ltd pp 281-283 at paras 85-88
[60] [2011] 3 SCR 269
[61] [2011] 1 WLR 1743 (see paragraphs 92-95 below)
[62] Fevaworks Solutions Ltd p 273 at para 54
[63] at para 6.29 at p 225
[64] at para 6.27 at pp 221-222
[65] at pp 291-292 and 294-295
[66] [2012] 3 CMLR 6 at p 104
[67] at p 116
[68] “[if] it were to respond to every complaint by requiring the offending material to be taken down, it would be making significant inroads into freedom of expression” (at p 119)
[69] at pp 117-118
[70] at p 119
[71] at p 125
[72] at p 125
[73] at para 49 and footnote 44
[74] [2013] EMLR 14 at p 308
[75] at p 315
[76] at p 317
[77] at p 318
[78] which held that by not removing defamatory material from a notice board after notification it might be inferred that an individual had associated himself with, or made himself responsible for its continued presence, thereby becoming publisher of that material
[79] see Gatley para 6.29 at p 226 in which the editors consider this view to be correct and more consistent with authority
[80] see paragraphs 72-75 above
[81] [2013] NZHC 540 (25 october 2012)
[82] at paras 116-117
[83] at para 53 and footnote 45
[84] at para 6.29 at p 226
[85] at p 1757
[86] at p 1758
[87] at 1758
[88] [2010] EWHC 616 (QB) (23 March 2010)
[89] at para 73
[90] at para 74
[91] [2012] VSC 88 (15 March 2012)
[92] [2012] VSC 533 (12 November 2012)
[93] eg Bunt, Metropolitan International Schools Ltd, Tamiz, and Crookes
[94] at para 15
[95] [2012] NZHC 2352 (12 September 2012)
[96] see Altimo Holdings Investment Ltd & ors v Kyrgyz Mobil Tel Ltd & ors [2012] 1 WLR 1804, 1825-1826 (which case was concerned with an application to set aside service out of jurisdiction)
[97] [2008] 5 HKLRD 631, 640
[98] para 4.136 at footnote 227 at p 101
[99] [2013] FCA 60 (2 February 2013), not cited by counsel
[100] Clark 2nd Aff para 17
[101] 2011 BCSC 1196 (2 September 2011), noted cited by counsel
[102] 2014 BCSC 1063 (13 June 2014)
[103] 2014 BCCA 295 (23 July 2014)
[104] C-131/12 [not yet published]
[105] but on 30 July 2014 the House of Lords European Union Committee’s Sub-committee on Home Affairs, Health and Education published EU Data Protection law: a ‘right to be forgotten’? and concluded that the Directive of the European Parliament and of the Council on 24 October 1995 on data protection and the interpretation of the Directive by the Court of Justice of the European Union do not reflect the current state of communications service provision where global access to detailed personal information has become part of the way of life, that the proposed “right to be forgotten” or “right to erasure” is misguided in principle and unworkable in practice, and that there are strong arguments for saying that search engines should not be classed as data controllers
[106] see Loutchansky v Times Newspapers Ltd & ors (Nos 2-5) [2002] QB 783, 817
[107] [2009] ECHR 451
[108] at para 42
[109] [2007] 1 AC 359
[110] at pp 288-289
[111] [2011] 3 SCR 269, 289-290
[112] [2001] 2 AC 127, 238
[113] at pp 289-290 (see also Chau Hoi Shuen, Solina Holly at para 5.13)
[114] in that they mean or are understood to mean inter alia that Yeung is a member of or associated with the triad, that he is a pervert, that he is a member of or associated with Sun Yee On, a notorious triad society in Hong Kong, that he has engaged in criminal activities through the triad, and that he is corrupt, dishonourable and immoral
[115] see Gatley para 11.4 at pp 395-396
[116] at p 966
[117] see Ratcliffe v Evans [1892] 2 QB 524, 528 and English and Scottish Co-operative Properties Mortgage and Investment Society, Limited v Odhams Press, Limited [1940] 1 K B 440, 461
[118] para 32.52 at pp 1256-1257
[119] CACV212/2007 (unreported, 13 February 2008)
[120] see para 23
[121] HCA378/2010, Sakhrani J (unreported, 20 May 2011)
[122] at pp 963-964
[123] [1937] 1 All ER 725
[124] [1999] EMLR 724
[125] at pp 964-965
[126] HCA12519/1997, Findlay J (unreported, 18 May 1999)
[127] at para 21
[128] Google AdWords is an online platform for pay advertising service operated by Google Inc that allows
hoqrvw}~”?ž º»ÄÆÈÍÏÙÚäåæûüý 3 4 E J ÷î÷î÷äÛÒ÷îÉ÷î÷Ò÷À²¦˜Û˜?À††Ûvmd÷d registered users to promote their business, raise awareness and increase traffic to their websites for a prescribed fee payable to Google Inc, and in summary the operation of such service is for the user to select a number of “keywords” for his advertisement campaign so that whenever an internet user “googles” by using any of the selected “keywords” the user’s website may appear at a prominent location of the search result page
[129] Google Trends is a tool provided by Google Inc to show the popularity of a particular search keyword in a specified location at a certain time by analysising a portion of the “google” web searches to compute how many searches have been done for the keyword entered relative to the total number of searches done on Google Search over time, which analysis indicates the likelihood of a random user to search for a particular search keyword from a certain location at a certain time, and the results of a user’s check on the “popularity” of the selected keyword will be shown on a graph, plotted on a scale from 0 to 100
[130] ie the approximate 12-month average of the number of “google” searches for such keywords
[131] see Leung Kin Fook & ors v Eastern Worldwide Co Ltd [1991] 1 HKC 55, 58-59, and also Enertec Co, Ltd v Gold Hill Hong Kong Holdings Limited HCA 2328/2012, Ng J (unreported, 21 January 2014)
[132] see In re J L Young Manufacturing Company, Limited [1900] 2 Ch 753 and Lumley v Osborne [1901] 1 QB 532
[133] [1994] 1 HKC 494
[134] [1991] 1 HKC 55
[135] see p 500
[136] see Bryne v Deane [1937] 1 KB 818, 837
[137] see Jameel (Yousef), Mohammed Hussien Al Amoudi v Jean Charles Brisard & anor [2006] EWHC 1062 (QB) (12 May 2006), Brady v Norman [2008] EWHC 2481 (QB) (20 October 2008) and Carrie v Tolkien [2009] EWHC 29 (QB)
[138] see Collins para 4.09 at p 71
[139] [2001] EMLR 876
[140] see Gatley para 32.9 at p 1228
[141] [2005] EWCA Civ 288 (3 March 2005)
[142] [2008] EWHC 658 (QB) (10 March 2008)
[143] [2008] EWHC 3155 (17 December 2008)
[144] at para 15
[145] at para 36
[146] see Collins (2014) para 4.14 at p 73 and ZAM v CFW & anor [2013] EWHC 662 (QB) (26 March 2013) paras 109 and 111
[147] para 11/4/3C at pp 176-177
[148] HCA1807/2011 (unreported, 5 November 2013) paras 110-119 (see also Wo Fung Paper Making Factory Ltd v Sappi Kraft (Pty) Ltd [1988] 2 HKLR 346, 357-358 and Prediwave Corp & anor v New World TMT Ltd [2007] 4 HKC 207, 221-222)
[149] see Hong Kong Civil Procedure 2014 para 11/4/4 at p 178
[150] see Hong Kong Civil Procedure 2014 para 11/4/4A at pp 178-179
[151] see paragraph 10 above
[152] at p 1263
[153] [2012] EWHC 449 (QB)
-----------------------
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
U
V
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
U
V
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
U
V
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
U
V
................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.
Related download
Related searches
- 2004 f150 5.4 belt diagram
- 2004 facts and events
- 2004 ford 4.6 engine problems
- 2004 f 150 engine problems
- 2004 ford f 150 problems
- ford f150 2004 engine problems
- 2004 ford 4 6 engine problems
- 2004 f150 5 4 crate engine
- 2004 ford 4 0 engine diagram
- 2004 f150 5 4 belt diagram
- 2004 f150 5 4 engine
- 2004 f150 5 4 replacement motor