Duffy v Google LLC
[2023] SASC 13
•3 February 2023
Supreme Court of South Australia
(Civil)
DUFFY v GOOGLE LLC
[2023] SASC 13
Judgment of the Honourable Auxiliary Justice Tilmouth
DEFAMATION - ACTIONS FOR DEFAMATION - PARTICULARS - OF STATEMENT OF CLAIM OR DECLARATION - PUBLICATION
DEFAMATION - ACTIONS FOR DEFAMATION - PLEADING - SOUTH AUSTRALIA
DEFAMATION - ACTIONS FOR DEFAMATION - TRIAL - FUNCTIONS OF JUDGE AND JURY - IN GENERAL
DEFAMATION - OTHER DEFENCES - TRIVIALITY
DEFAMATION - OTHER DEFENCES - MISCELLANEOUS DEFENCES
DEFAMATION - PUBLICATION - GENERALLY - INTERNET PUBLICATIONS
The applicant has brought an action for defamation against the respondent. The applicant alleges that the respondent published material defamatory of her by way of providing as search results on google.com and google.com.au two URLs to a third-party website, referred to as RoR1 and RoR2, and three associated search result snippets.
The respondent submits that the applicant did not make out all the elements of defamation as there was insufficient proof of publication and because the snippets, when considered in the context of the wider search results, did not amount of defamation. The respondent further relies on the defence of innocent dissemination, arguing that the respondent was not required to remove the URLs because the applicant did not provide proper notice. The respondent further contends that the matter is trivial.
Held:
1.Both RoR1 and RoR2, as well as all three snippets, on google.com.au were defamatory and were published by inference to unknown third parties.
2.There was no sufficient evidence that RoR1 or RoR2 were accessed on google.com in Australia and hence it cannot be accepted on the balance of probabilities that they were published in Australia.
3.Google is proven to have participated in the communication of RoR1 and RoR2 in Australia on google.com.au so as to render it liable as a secondary publisher.
4.The defences of innocent dissemination and triviality fail.
5.The parties are to be heard on the issue of consequential damages, before questions of costs arise.
Communications Decency Act 1996 (US) s 230; Limitation of Actions Act 1936 (SA) ss 37(1), 37(2); Uniform Civil Rules 2020 (SA) r 69.2; Defamation Act 2005 (SA) ) ss 10A, 12A(1)(a)(ii), 14(2), 14(3), 21, 31; Evidence Act 1929 (SA) s 34C(1a); Juries Act 1927 (SA) ss 5, 11; Juries Act (No 1 of 1862) (SA) s 26; Juries Act 1917 (SA) s 134; Criminal Code Act 1889 (Qld) s 359B(c)(ii); Criminal Code Act 1924 (Tas) s 192(1)(g)-(h); Criminal Law Consolidation Act 1935 (SA) s 19AA(1)(a)(iva)‑(ivb); Crimes Act 1958 (Vic) s 21A(2)(ba); Crimes Act 1900 (ACT) s 35(2)(h); Defamation (Miscellaneous) Amendment Act 2020 (SA) sch 1 cl 2, referred to.
Trkulja v Google LLC (2018) 263 CLR 149, applied.
Google LLC v Defteros (2022) 403 ALR 434, distinguished.
Duffy v Google Inc (2015) 125 SASR 437; Google Inc v Duffy (2017) 129 SASR 304; Duffy v Google LLC [2019] SASC 157, discussed.Google Inc v Trkulja (2016) 342 ALR 504; John Giordano v Romeo (Fla 3rd Dist Ct App, No 3D11‑707, 28 December 2011); Duffy v Google Inc (No 3) [2016] SASC 1; Duffy v Google Inc (No 2) [2015] SASC 206; Dasreef v Hawchar (2011) 243 CLR 588; Dow Jones & Co Inc v Gutnick (2002) 210 CLR 575; Ostrowe v Lee (1931) 175 NE 505; Brunswick, Duke of v Harmer (1849) 14 QB 185; Sands v Channel Seven Adelaide Pty Ltd (2009) 104 SASR 452; Al Amoudi v Brisard [2007] 1 WLR 113; Capolingua v Nationwide News Pty Ltd [2016] WASC 156; Dods v McDonald (No 1) [2016] VSC 200; Stoltenberg v Bolton (2020) 380 ALR 145; David v Abdishou [2012] NSWCA 109; Trkulja v Yahoo! Inc LLC [2012] VSC 88; Blatch v Archer (1774) 98 ER 969; Jones v Skelton [1963] 1 WLR 1362; Favell v Queensland Newspapers Pty Ltd (2005) 221 ALR 186; Mirror Newspapers Ltd v Harrison (1982) 149 CLR 293; Chakravarti v Advertiser Newspapers Ltd (1998) 193 CLR 519; Lewis v Daily Telegraph [1964] AC 234; Lord William Nevill v The Fine Art and General Insurance Company, Limited [1897] AC 68; Farquhar v Bottom [1980] 2 NSWLR 380; John Fairfax Publications Pty Ltd v Rivkin (2003) 201 ALR 77; Morosi v Broadcasting Station 2GB Pty Ltd [1980] 2 NSWLR 418; Crookes v Newton [2011] 3 SCR 269; Byrne v Deane [1937] 1 KB 818; Fairfax Media Publications Pty Ltd v Voller (2021) 95 ALJR 767; Thompson v Australian Capital Television Pty Ltd (1996) 186 CLR 574; Defteros v Google LLC [2020] VSCA 219; McNicol v Grandy [1932] 1 DLR 225; Barrow v Bolt [2015] VSCA 107; Chappell v Mirror Newspapers Ltd (1984) Aust Torts Rep 80-691; Morosi v Mirror Newspapers Ltd [1977] 2 NSWLR 749; Jones v Sutton (2004) 61 NSWLR 614; Associated Newspapers Ltd v Dingle [1964] AC 371; Hockey v Fairfax Media Publications Pty Ltd (2015) 237 FCR 33; Humphries v TWT Ltd (1993) 120 ALR 693; Search King v Google Technology Inc. 2003 WL 21464568 (W.D. Okla. 2003); Metropolitan International of Schools Ltd t/as Skillstrain and/or Train2Game v Designtechnica Corp t/as Digital Trends [2011] 1 WLR 1743 (QB), considered.
DUFFY v GOOGLE LLC
[2023] SASC 13Civil
TILMOUTH AJ.
Overview of proceedings
The parties
Background in brief
The nature of internet searches
Causes of action
Ripoff Reports
Previous proceedings
Preliminary issues
The tort of defamationPublication
Legal principles
Direct evidence
Inferential evidence
google.com.au
google.com
Capacity to convey defamatory imputations
The snippets
Defamatory meanings?
Enticement or encouragement
The Ripoff Reports
Innocent dissemination
Legal principles
Google not coming within the rule
Google neither knew nor ought reasonably to have known of defamatory characterGoogle as an innocent disseminator?
Triviality defence
Miscellaneous
Conclusion
Overview of proceedings
The within action is brought by Dr Janice Duffy in the tort of defamation claiming damages for allegedly defamatory publications by Google Inc through its search engines on the World Wide Web.[1] In essence her case is that Google defamed her by publishing highly offensive third-party webpages by means of indexing hyperlinks leading to defamatory websites presented as results on its search engines. She claims the defamatory content remained indexed on the first page of Google search results on domains available in Australia for upwards of eight years.
[1] Google Inc is hereafter referred to as “Google”.
Google denies liability, contending that the subject search results were not defamatory, that they were not published to any third party and it calls in aid the defences of innocent dissemination and triviality. Although pleaded, the defence of qualified privilege was abandoned during the course of the hearing.[2] An order was made on 25 August 2021 for the trial to proceed first on the issue of liability leaving the issue of damages for later consideration.
[2] T187.22-25.
The parties
Dr Duffy completed Honours and Masters degrees majoring in sociology and politics before she was awarded a PhD from Flinders University in 1999 for her thesis on the topic of alcohol in the workplace. She was until August 2010 a senior researcher in the fields of public health and clinical medicine within the South Australian Health Department, before resigning and becoming unemployed.
Google is a public company incorporated under the laws of the State of Delaware in the United States of America. It is the owner and operator of a global internet business, including internet search engines. Before October 2017 it was the owner and operator of many websites with Country specific domains. It owned and operated google.com.au in Australia and google.com in the United States. Google then transitioned to the use of ‘domainless’ Country editions so that, for example, users detected as within Australia were by default served with the Australian edition of the Google search engine even if a browser address bar displayed google.com.[3] On these sites it hosts a search engine tool available to the public at large for searching the World Wide Web.
[3] Exhibit A1A, [62]-[65]; Exhibit R7A, [9]-[10]; T687.9-22, T692.23-31.
Background in brief
At some time in 2006 Dr Duffy began searching the website Kasamba which offered counselling as well as spiritual advice. By mid-2007 she began to find a large number of complaints about this website posted on another website RipoffReport.com.
Ripoff Report is owned by a US-based entity totally unrelated to Google. Its site operates as an interactive forum for users to create and post online mostly derisory “reports” and to allow users to add comments to such reports. Between mid-2007 and early 2009 she posted either 14 or 16 complaints on RipoffReport.com.[4] According to her, those posts drew some support and some criticism, so she formed a support group with others “vulnerable” to defamatory posts.[5] As time went on, she became aware of posts aiming to discredit her and attempting to discourage members of the support group. She further discovered that there were adverse posts about her and her support group on Ripoff Report, appearing in search engine results when searching her name.
[4] T220.13-221.23; Closing Written Submission of Applicant, [7].
[5] T221.13-222.8.
Between 7 and 12 September 2009, she communicated with Google Australia Pty Ltd, Google Inc, [email protected], Google removals team, and the Google legal team in an attempt to prevent Ripoff Reports appearing in searches for her name on Google’s search engine.[6] On 14 September 2009, Google informed Dr Duffy by email that it required her to identify which Country‑specific site was involved, what search terms were used in addition to the precise uniform resource locator (“URL”) complained of before it would consider removal.[7]
[6] Master Chronology, pp 1-2.
[7] Exhibit R1.
Over the course of further communications during 2009 and 2010, Google consistently refused to remove any URLs at the request of Dr Duffy. In the meantime, Dr Duffy issued proceedings in her first action on 16 February 2011.[8] Between 14 April 2011 and 3 December 2012, there was a series of correspondence between counsel for Dr Duffy and counsel for Google concerning removals.[9] Google eventually removed a number of offending Ripoff Report URLs relating to Dr Duffy from appearing in search results obtained on google.com.au between 7 March and 29 November 2011 as follows:
·five on 7 March 2011;
·16 on 31 May 2011;
·14 on 2 November 2011;
·15 around 14 November 2011; and
·three on 29 November 2011.[10]
Thereafter it appears that Dr Duffy took no further steps to secure blocking by Google between 3 December 2012 and 28 April 2014.
[8] In the District Court of South Australia, later transferred to the Supreme Court of South Australia, hereafter mostly referred to as the ‘first action’.
[9] Exhibit R1.
[10] Exhibit R7B.
All the same Ripoff Report URLs remained published on google.com which were able to be accessed in Australia once the user manually changed from the google.com.au search engine to the google.com search engine. Google made it clear that it would not block those search results appearing on google.com, which it considered to be protected in the United States by s 230 of the Communications Decency Act 1996, which serves to immunise online publishers against suits for defamatory content the publisher did not create.[11] Whatever the exact situation, Google maintained a settled policy of refusing to remove material from google.com based on the Constitutional protection of freedom of speech in the United States.
[11] Exhibit A1A, [14].
The present action in the Supreme Court was issued on 19 October 2016 following a judgment in her favour in the first action against Google for defamation delivered on 27 October 2015.[12]
[12] FDN 1 and 2; Duffy v Google Inc (2015) 125 SASR 437, (Blue J).
Soon after the judgment in the first action, Dr Duffy became aware that practically identical search results were once again available on Google’s search engines.
Be that as it may, Google continued to insist on notification of complete URLs of concern to Dr Duffy which were appearing in search results on google.com.au before it would consider removal. On 5 November 2015, Dr Duffy filed her seventh affidavit in the first action which annexed a proposed fourth amended statement of claim which particularised a URL she referred to as the first search result URL. Around 5 November 2015, Google removed the first search result URL. On 9 November 2015 it extended an open invitation to Dr Duffy to notify it of any other URLs appearing in search results on google.com.au on that understanding.[13] Between mid-December 2017 and early July 2018, around a further 11 URLs were blocked by Google. All told Google blocked 64 URLs between 7 March 2011 and 9 July 2018. Only 13 were removed after 27 October 2015, the first two of which were on 5 November 2015, two more on 18 December 2017, one on 7 March 2018, one on 8 July 2018 and the remaining seven on 9 July 2018.[14]
[13] Exhibit R8 Volume 2, Tab 6, p 672.
[14] Exhibit R7A and R7B.
Over the course of time following a succession of interlocutory proceedings, Dr Duffy finally filed the third amended statement of claim on 29 July 2022. In sum, this action originally came down to four specifically pleaded URLs containing allegedly defamatory material. This was reduced to two as a result of concessions made by her during the course of the trial following protracted submissions in relation to time limitation issues.[15]
[15] T762.8-771.15, particularly T770.23-771.11.
This action is based on two URLs (RoR1 and RoR2) and two search engines (google.com.au and google.com). RoR1 was allegedly published on google.com.au between 19 October 2015 and 6 November 2015. RoR1 was also allegedly published on google.com between 19 October 2015 and 19 October 2016. RoR2 was allegedly published on google.com.au between 19 October 2015 and 19 October 2016. Finally, RoR2 was also allegedly published on google.com between 19 October 2015 and 19 October 2016.
The nature of internet searches
The nature of an electronic search enquiry made on computers by means of a search engine is now so well documented that it is unnecessary to repeat much of what appears in decisions of Australian courts.[16] It is sufficient for present purposes to simply observe the salient features:
·The internet is a system of globally linked computer networks collectively referred to as the World Wide Web communicating by means of a standardised protocol regularly used by Australians;
·It is comprised of a vast system of constantly changing and expanding webpages generated by millions of people and organisations worldwide using so called HyperText Markup Language (HTML);
·Each webpage has a unique URL, which acts as the internet address for that webpage;
·The HTML code of a webpage is read by a software program known as a browser, which displays text and images on the user’s device;
·A webpage typically contains hyperlinks to other webpages which, when selected, the browser displays text and images to that other webpage, thus leaving the first webpage; and
·A search engine is an automated information retrieval system designed to allow the user to navigate and search the Web by the use of user‑entered queries.
[16] See Google LLC v Defteros (2022) 403 ALR 434, [10]-[11], [78]-[85], [117]-[125]; Google Inc v Trkulja (2016) 342 ALR 504, [159], [157], [170]-[171], [180]-[182], [192], [194], [199], [202].
Specifically with respect to Google, the following propositions emerge from the cases:
·Over 100 billion searches are made by users on Google’s search engine worldwide each month;
·In order to catalogue the trillions of pages constituting the Web so as to provide results for search queries, Google operates a Web crawler program which identifies webpages available on the Web;
·Google’s search engine uses an indexing program to organise the data obtained by the Web crawler into a form that is more easily searched by computer algorithms by building a list of every webpage that contains each search word;
·When a user enters a search query into Google’s search engine, it conducts a Web search almost instantaneously by means of a fully automated process without human intervention, with an averaged response time of about half a second;
·Google’s search engine uses sophisticated and confidential computer algorithms to make predictions about which indexed webpages are most likely to be of interest to a user and in which order in response to the user’s search query;
·The search engine then returns a list of results in response to a search query in the form of snippets containing hyperlinks to third party webpages;
·The order of snippets is determined by an assessment of such factors as how frequently or recently the webpage is updated, the popularity of the website, where the subject matter of the webpage is geographically located, how many requests it can handle in a given timeframe, the number of times the user’s search terms appear on the webpage, how often other webpages link to that webpage, the importance of the linking webpages and the user’s previous search history;
·Google’s algorithm performs a search function afresh every time a user enters a new search so that the data produced by the ranking algorithm will never be the same twice; and
·Google can prevent a specified webpage from crawling and/or indexing and/or having a snippet displayed or a cached copy made available to a user undertaking a search on Google websites, but this does not remove the webpage from the Web; it can still be accessed directly or found using other search engines.
Causes of action
In support of her case, Dr Duffy undertook a series of searches of various terms in and about her name on both google.com.au and google.com over time. The hard copy documents evidencing these searches were tendered in the course of the trial. The manner of proof was somewhat confusing and disjointed, which is unsurprising given that she was unrepresented. Matters compounded with the submission by Dr Duffy of six volumes of trial books and eight volumes of tender books, making it difficult and time consuming to cross reference with each other and with Google’s three volumes of tender books.[17]
[17] Exhibits A2, A3 and R8 respectively.
The first search result URL complained of is demonstrated by the result of a search by Dr Duffy using the term “janice duffy” on google.com.au on 29 September 2015, the first search result of which produced the following snippet (first snippet):[18]
Ripoff Report | Janice Duffy - Psychic Stalker! Psychics …
16, 2014 - Janice Duffy - Psychic Stalker! Psychics Beware Of Australian Psychic Stalker! Janice Duffy Stalked me on the computer for several months.[18] Applicant’s Trial Book (ATrB) V1, p 18; Closing Written Submission of Applicant, [31].
The first search result for the search term “dr janice duffy” also conducted by Dr Duffy on 29 September 2015 is the same Ripoff Report, but a different snippet of the page is shown (second snippet):[19]
Ripoff Report | Janice Duffy - Psychic Stalker! Psychics…
16, 2014 - *UPDATE Employee: Dr Janice Duffy Psychic Stalker Repatriation General Hospital Administrator Extorted. *UPDATE EX-employee responds: …[19] ATrB V1, p 21.
These snippets contained a hyperlink opening to the RipoffReport.com webpage contained in Annexure 2 bearing the number 295712.[20] This Ripoff Report was referred to as RoR1. It was first published on RipoffReport.com on 30 December 2007[21] and the copy in Annexure 2 was accessed on 2 November 2015 by Dr Duffy.[22]
[20] ATrB V1, pp 34-37.
[21] Exhibit R8, p 456.
[22] ATrB V1, pp 34-37 plus a blank page between pp 37 and 38.
The dates of publication relied upon with respect to these snippets and this Ripoff Report, between 19 October 2015 and 6 November 2015, derive from screenshots of six searches on google.com.au using the search terms “janice duffy” and “dr janice duffy” between 23 October 2015 and 6 November 2015 contained in Annexure 1 to her third amended statement of claim.[23] Every screenshot in Annexure 1 displays RoR1 as the first search result. This date range is further informed by the fact that by 7 November 2015 Google removed this offending URL from google.com.au, but not from google.com.
[23] ATrB V1, pp 18-33.
With respect to google.com, the snippets and the URL are the same as with google.com.au. The alleged publication date range of 19 October 2015 to 19 October 2016 is supported by a series of search results in Annexure 4 of the third amended statement of claim regarding RoR1. Searches were conducted for both “janice duffy” and “dr janice duffy” on 29 September 2015, for both on 23 October 2015, for both on 7 and 23 November 2015, for both on 15 December 2015, for “dr janice duffy” only on 18 December 2015, for both on 22 December 2015, for both on 17 January 2016, for “dr janice duffy” only on 9 and 13 February 2016, for both on 1 March 2016, for both on 8 May 2016, for both on 16 and 19 June 2016, for “dr janice duffy” only on 21 June 2016, for both on 22 July 2016, for “janice duffy” only on 24 August 2016, for “janice duffy” only on 5 November 2016, for “janice duffy” only on 21 November 2016, for both on 3 January 2017, for both on 14 January 2017 and for both on 28 January 2017.[24] The searches on 5 and 21 November 2016 were conducted using the Firefox browser. All other searches were completed on the Google Chrome browser.
[24] ATrB V1, pp 67, 69, 71, 73, 75, 77, 78, 80, 83, 85, 87, 89, 91, 93, 95, 97, 99, 101, 103, 105, 108, 110, 112, 114, 116, 117, 118, 120, 121, 124, 127, 129, 131, 133, 135, 136 and 138, respectively.
The second pleaded URL was demonstrated by the result of a search once again undertaken by Dr Duffy on google.com.au using the term “janice duffy reviews” on 6 November 2015, the third search result of which produced the following snippet (third snippet):[25]
Ripoff Report | Dr. Janice Duffy Complaint Review
6, 2008 - Dr. Janice Duffy Complaint Review: Dr. Janice Duffy Stop the Australian Psychic Stalker Dr. Janice Duffy!! Adelaide, South Australia Adelaide,…
[25] ATrB V1, p 173.
This snippet contained a hyperlink opening to the RipoffReport.com webpage contained in Annexure 7 bearing the number 295925.[26] This Ripoff Report was referred to as RoR2. It was first published on RipoffReport.com on 31 December 2007,[27] and the copy in Annexure 7 was accessed on 27 December 2015 by Dr Duffy.[28]
[26] ATrB V1, pp 194-198.
[27] Exhibit R8, p 465.
[28] ATrB V1, pp 195-198.
The pleaded dates of publication on google.com.au for this second URL of between 19 October 2015 and 19 October 2016 derive from seven further searches by Dr Duffy using the search term “janice duffy reviews” between 6 November 2015 and 30 May 2016 showing RoR2 as contained in Annexure 6.[29] Those searches were undertaken by Dr Duffy on 6 November 2015, on 8 May 2016, twice on 9 May 2016 (one using the Firefox browser), and on 18, 25, and 30 May 2016.[30]
[29] ATrB V1, pp 176-189.
[30] ATrB V1, pp 176, 179, 181, 184, 186, 187, 189 and 191 respectively.
The snippet, URL and pleaded dates of publication are the same for RoR2 on google.com. The search results are contained in Annexure 8. Dr Duffy conducted searches on 27 December 2015, 9 and 13 February 2016, 8 May 2016 (using the Firefox browser), 18 May 2016 (using the Firefox browser), 30 May 2016, and 16 and 19 June 2016.[31]
[31] ATrB V1, pp 200, 204, 206, 208, 205 (note after p 209 there is a repetition in numbering from 205 to 209), 206, 207, 208, 210, and 211 respectively.
All these dates derive in part from the primary requirement contained in s 37(1) of the Limitation of Actions Act 1936 (SA) that a cause of action for defamation must be brought “after the end of a limitation period of 1 year running from the date of the publication”. The present action was lodged on 19 October 2016, that is approximately one year after judgment was given in the first action. Subsection 37(2) of the Limitation of Actions Act, as it stood at relevant times, compelled the court to extend the limitation period to up to three years if satisfied that “it was not reasonable in the circumstances for the plaintiff to have commenced an action in relation to the matter complained of within 1 year from the date of the publication”.[32] Dr Duffy’s attention was drawn to this provision as well as to r 69.2 of the Uniform Civil Rules 2020 (SA) which permits an amendment in the discretion of the court “if the new cause of action arises out of substantially the same facts as the original cause of action”. Despite all this, Dr Duffy did not press for any extension of time under either avenue of redress.[33]
[32] Subsection 37(2) was substituted with a new requirement for time extensions which came into effect on 1 July 2021: Defamation (Miscellaneous) Amendment Act 2020 (SA) sch 1 cl 2.
[33] T770.11-771.26.
It follows that any action was statute barred by 19 October 2016 in the case of both Ripoff Reports. The result is that no claim of publication prior to 19 October 2015 or after 19 October 2016 is maintainable because of the limitation period. Dr Duffy introduced two further instances of publication in 2017 and 2018 in her second amended statement of claim, filed on 10 March 2020, and her third amended statement of claim, filed on 29 July 2022. Absent valid orders for an extension of time, these two further pleaded acts of publication are statute barred since they were first introduced more than one year after publication. Hence there is no applicable time limitation exemption so the otherwise complex related limitation issues have dissolved owing to the concessions made by Dr Duffy referred to above.
All of these searches, reduced to hard copy format, were reproduced in the several Annexures submitted by Dr Duffy. Those only consist of the first page of Google search results which shows a set of 10 search snippets on one page. This can be seen by the emboldened 1 under the first “o” in the “Goooooooooogle” Logo at the centre bottom of each page. For each search query, there were potentially hundreds of thousands of search results found.
The gist of Dr Duffy’s complaints arise from the fact that all of the above searches contained a hyperlink which, once selected, opened to the Ripoff Report webpage containing material defamatory of her.
Ripoff Reports
These Ripoff Reports were created on a website operated by Xcentric Ventures LLC located at ripoffreport.com.[34] The website Ripoff Report purports to be a consumer review forum for those believing to be the victim of unfair or illegal business practices. In truth, it profits from extortive business practices. It has a vast database of damaging content about individuals and organisations which as a matter of fixed policy it will not remove. Those adversely affected can be charged substantial fees in order to have Ripoff change a report, alter webpages or display more favourable content.
[34] Exhibit A1A, [17].
The expert witness Mr Silver-Smith called by Dr Duffy described Ripoff Report as:[35]
… something of a magnet for people that had a complaint or wished to report on something but it also became a place where people posted a lot of negative reviews or negative opinions about individuals, and in some cases they also posted fraudulent information, information that was untrue about individuals and companies and their interactions with them.
[35] T556.28-35.
In his report he described its website as one of the “more predatory practices … taking advantage of Section 230 protections” and “likely the most notorious”.[36] He added under cross-examination:[37]
I did say that, and I think that that's readily apparent that if you're providing a positive review about a company and Ripoff Report comes up in the search results, and you see company name Ripoff Report in the title in search results, it's going to look negative regardless of the content of that review, but because of the nature of the site it became fairly pernicious and was a magnet for all sorts of negative content.
[36] Exhibit A1A, [16].
[37] T579.17-579.25.
The Third District Court of Appeal for the State of Florida said of Xcentric:[38]
The business practices of Xcentric, as presented by the evidence before this Court, are appalling. Xcentric appears to pride itself on having created a forum for defamation. No checks are in place to ensure that only reliable information is publicized. Xcentric retains no general counsel to determine whether its users are availing themselves of its services for the purpose of tortious or illegal conduct. Even when, as here, a user regrets what she has posted and takes every effort to retract it, Xcentric refuses to allow it. Moreover, Xcentric insists in its brief that its policy is never to remove a post.
[38] John Giordano v Romeo (Fla 3rd Dist Ct App, No 3D11-707, 28 December 2011), p 4.
The evidence was that if Google successfully removed a Ripoff Report URL containing defamatory content, it subsequently reappeared in Google search engines because Ripoff was in the habit of changing URLs specifically to avoid Google’s removal methods.[39]
[39] Exhibit A1A, [26]-[27]; T492.5-13, T557.28-37.
Previous proceedings
Dr Duffy brought the first action in defamation against Google in February 2011 in the District Court. In 2014, the action was transferred to the Supreme Court. Dr Duffy eventually went to trial alleging publication of several Ripoff Reports and several other websites that published material ostensibly derived from the Ripoff Reports. In October 2015, Blue J published reasons in which he concluded Dr Duffy had proven Google published several webpages in 2010 containing various defamatory imputations.[40] He later assessed damages at $100,000 and in January 2016 awarded judgment in favour of Dr Duffy against Google for $115,000 inclusive of interest for extracts from three webpages and the two ripoffreport.com webpages.[41]
[40] Duffy v Google Inc (2015) 125 SASR 437, [474]-[481].
[41] Duffy v Google Inc(No 3) [2016] SASC 1, [2]-[3].
On 5 November 2015, Dr Duffy filed an application in the first action seeking to amend her statement of claim to add a plea of aggravated damages based on further publications in 2014 and 2015 of two similar Ripoff Report URLs.[42] However Blue J considered that “any publication in 2014-2015 by reference to a different URL gives rise to a separate and independent cause of action” and consequently that “Dr Duffy is free to institute a separate action for publication in 2014-2015”.[43] Dr Duffy’s solicitors then filed on 6 September 2016 an amended interlocutory application in the first action, seeking permission to file a proposed statement of claim in what ultimately became these proceedings.[44] She obtained leave to proceed pursuant to s 21 of the Defamation Act 2005 (SA) on 8 September 2016.[45]
[42] Exhibit R8, p 682, [13].
[43] Duffy v Google Inc (No 2) [2015] SASC 206, [37].
[44] Exhibit R5.
[45] Exhibit R5.
An appeal by Google to the Full Court in the first action was dismissed on 4 October 2017.[46] Later Blue J was called upon to consider an application by Dr Duffy to amend her statement of claim and to proceed with the action pursuant to s 21 of the Defamation Act and an application by Google for summary judgment. His Honour dismissed Google’s application for summary judgment and granted both of Dr Duffy’s applications.[47]
[46] Google Inc v Duffy (2017) 129 SASR 304.
[47] Duffy v Google LLC [2019] SASC 157, [151]-[155].
Preliminary issues
Dr Duffy issued a subpoena for the attendance of the potential witness Professor Hennessy who was until 2016 the president of Stanford University and was at the time of trial Chairman of Google’s holding company, Alphabet Inc.[48] Dr Duffy sent a number of emails to him over the course of time, including whilst the first action was in progress as well as afterwards. These contained complaints about the offending webpages numerous times, yet Google refused to remove or block access to them when she maintains it should have. She proposed to ask him, if called to give evidence, why that was the case. During a directions hearing on 25 August 2022, the Chief Justice set aside the subpoena on the basis it was admitted by Senior Counsel for Google that it, through Professor Hennessy, was put on notice of the claimed defamations and it was then a matter for Google to determine whether to adduce evidence in its own case explaining why nothing was done.
[48] FDN 126.
The application for leave to issue the subpoena was renewed at the beginning of this trial and refused for similar reasons. Quite apart from the inherent problem of leading as opposed to cross-examining Professor Hennessy, as the onus is on Google to justify its inaction if it can, it remained a forensic decision for it to make to call him or not. In any event Google made admissions in response to a Notice to Admit issued by Dr Duffy that went somewhat further than the admissions made before the Chief Justice, in as much as there were now concessions that the email from Dr Duffy of 3 October 2016 was “read either by Prof Hennessy or on his behalf” and that he “either had actual knowledge of the content of the email, or constructive knowledge”.[49]
[49] FDN 151, item 43; T175.28-176.27.
The Court further determined the admission into evidence of two expert reports obtained at the instance of Dr Duffy. The first was by Mr Silver-Smith dated 29 June 2022, provided in his capacity as “an expert and consultant in a number of areas, including Internet marketing, social media, online reputation management … search engine optimisation (“SEO”), Internet advertising and related practices”.[50] It is convenient to focus for the present purpose on the summary provided by Mr Silver‑Smith in his report, condensed as follows:[51]
IX SUMMARY OPINIONS
Google’s arguments that they could not reasonably deduce the proper URLs of pages submitted for removals if they only received elided URLs appears to be completely facile, if the provided elided URLs included the website’s domain name. I found that with the elided URLs shown in the screengrabs of Ripoff Report listings Dr. Duffy provided, there was sufficient information to rapidly and easily identify the pages in question at ripoffreport.com by using Google’s search operators.
…
Google could have fairly easily created a method to detect Ripoff Report’s subversion, and automatically counteracted it on behalf of defamation victims who otherwise would have had a reasonable expectation that their prior removal requests were continuing to protect their reputations from harm. As they apparently did not take action to counteract this, their inaction effectively allowed Dr. Duffy to be unnecessarily defamed further once more, along with all other people who had had Ripoff Report pages removed from Google’s search results over time.
In spite of claims to the contrary, Google can effect removals without requiring that all URLs of contravened content be provided to them. The have apparently developed such methods for similar types of content restrictions outside of the formal removal request system, and they announced in 2021 publicly that they would create a category of “known victim” so that they may automatically suppress defamatory content for some individuals.
This report also contained assertions expressed on moral grounds, calling for increased assistance for persons defamed or victimised online by third parties.
[50] Unredacted report of Mr Silver-Smith, [2].
[51] Exhibit A1A, [110], [112]-[113].
Objection was taken on the basis that Google’s evolving practices of seeking to maintain the integrity of its search algorithm was not capable of proving notice, knowledge or ostensible knowledge of the specific search results and URLs complained of, nor was it relevant to the legal question of publication, in that neither the aim nor purpose of Google in making information universally accessible is relevant to whether there is publication.[52]
[52] Google LLC v Defteros (2022) 403 ALR 434, [54].
The second expert report was that of Mr Zach Vorhies dated 7 July 2022.[53] This dealt with several topics which to summarise cover: asserted moral concerns and allegations of “censorship” by Google;[54] unproven allegations of historic use of manual overrides to the search algorithm in October 2017 with respect to a mass shooting event in May 2017;[55] alleged censorship of the translation of the word “covfefe” in response to a Donald Trump tweet;[56] inferences regarding the availability and use of manual overrides to the usual operation of Google’s search algorithm;[57] inferences regarding the evolution of Google’s search algorithm to take into account additional criteria in ranking search results, particularly for “Google News”;[58] an allegation that Google influenced or sought to influence the 2020 US election through amendments to its search algorithm particularly on “Google News”;[59] and steps Google could theoretically take to demote the appearance of webpages containing content having similar keywords to specified URLs in its search results, or to demote or remove an entire domain such as ripoffreport.com from its search results irrespective of its content.[60] Mr Vorhies was a software engineer employed by Google for over eight years. He worked with the “Google Earth” team for five and a half years before moving to YouTube for three years and resigning in June 2019.[61] During his time with Google he claims to have become aware of “Google’s vast censorship system”.[62]
[53] FDN 100.
[54] At [3]-[7].
[55] At [8]-[17].
[56] At [54]‑[59].
[57] At [18]-[32], [37]-[41].
[58] At [34]-[35], [43]-[46], [60]-[63], [87]-[89].
[59] At [64]-[86].
[60] At [90]-[98].
[61] FDN 100, [2]-[3].
[62] FDN 100, [3].
Objection was taken to this report on the same grounds as those pertaining to Mr Silver-Smith and compendiously as raising detailed technical and policy issues unrelated to the issues in dispute. It is further complained that he did not identify particular facts in issue to which the field of specialised knowledge or proposed opinion evidence is said to relate, did not articulate or identify the relevant field of specialised knowledge or expertise, expressed opinions based on hearsay or unproven material without clear articulation of the facts and assumptions on which those particular opinions were based, overstepped the bounds of his expertise and expressed value judgments, commentary or opinions on questions of policy.
Both reports were proffered by Dr Duffy as relevant to Google’s ability to remove content from its search results even without notice of the complete URL at which the content is hosted, to refute technical evidence previously provided by Google in the first action and to indicate a significant technical advancement of Google’s abilities to remove content through the provision of URLs.
As there was an “overarching objection” to the reception of both reports on the basis that they were irrelevant “to any pleaded issue” and not in any event intelligible as admissible expert opinion evidence,[63] it was necessary to rule on these objections either in advance of trial, or certainly before each witness was called.[64]
[63] FDN 134.
[64] Dasreef v Hawchar (2011) 243 CLR 588, [119], [123]-[124], [135].
The application to adduce this expert evidence in Dr Duffy’s case occurs in the background of the intimation by counsel for Google that it proposed to call its staff software engineer Dr Raghava Kondepudy. Rather than attempt to summarise it here, it is opportune to extract the summary prepared by Blue J in his judgment giving leave to proceed, on which the parties based their submissions in this action:[65]
Mr Kondepudy gave evidence that before October 2017 Google maintained different websites for different countries. The website google.com.au (the Australian domain) was maintained for the purpose of searches by users in Australia. The website google.com (the United States domain) was maintained for the purpose of searches by users in the United States of America. The algorithm for search results took into account the domain such that searches on the Australian domain would tend to be slanted toward results more relevant to Australian users. Removals or blocking also differed between different country domains. If a user in Australia (assessed primarily by reference to the user’s apparent IP address) accessed the United States domain (google.com), he or she would typically be redirected to the Australian domain but could override this default and undertake searches on the United States domain or any other country domain (although only a small percentage of users actually did so). If a user used a virtual private network (VPN) with the gateway to the Internet in a different country, Google would perceive the user as being in that different country.
Mr Kondepudy gave evidence that since October 2017 Google distinguishes between country editions: for example there is an Australian country edition and a United States country edition. By default, a user in Australia (assessed primarily by reference to the user’s apparent IP address) undertaking a Google search will typically be directed to the Australian country edition but this default position can be overridden by the user to access a different country edition.
Mr Kondepudy gave evidence in some detail concerning what Google variously terms its “removals” or “blocking” process. Google is able to block search results on one of its domains or country editions in respect of a specific URL. This means that a person searching on the relevant domain or country edition will not receive a search result which contains a hyperlink to that specific URL and hence will not be able to access the external webpage denoted by that URL via that Google domain or country edition. I refer, as does Google, to this as a removal. Obviously this does not entail removal of the external webpage and further a person using a different search engine (or a Google domain/country edition which has not been subject to the same removal) will still be able to access the external webpage. The removals process operates at the level of the external webpage’s URL. If the operator of the external website changes the URL in respect of the webpage, the removal will not operate in respect of the changed URL.
…
Mr Kondepudy gave evidence that it is Google’s legal team which decides whether a removal should be undertaken. If so, it instructs the Google software engineers to undertake the technical process of effecting the removal.
[65] Duffy v Google LLC [2019] SASC 157, [33]-[35], [37].
Following submissions on both sides, a redacted version of the report of Mr Silver-Smith was admitted as Exhibit A1A, for use on the subject of capacity to pulldown or remove URLs and particularly as to the issue of whether Google needed the entire URL to effect removal.[66] The report of Mr Vorhies was not admitted because it largely expressed unsupported or irrelevant opinions, contained moral or value judgments, or stated opinions which the Court was required to determine as final issues on questions of fact or conclusions of law.[67]
[66] T284.1-3, T545.11-550.19, T585.26-27.
[67] T287.15-289.20.
The tort of defamation
An action in defamation focuses upon publications causing damage to reputation.[68] Actionable harm to reputation occurs when a defamatory publication is comprehended by the reader.[69]
[68] Dow Jones & Co Inc v Gutnick (2002) 210 CLR 575, [25].
[69] Dow Jones & Co Inc v Gutnick (2002) 210 CLR 575, [26].
According to the written submissions of Google, of the matters in issue in this case, Dr Duffy bears the onus of proving to the civil standard of on balance of probabilities these elements of her cause of action in defamation:[70]
1.that Google published one or both matters now complained of in the specific manner pleaded;
2.the matter(s) proven conveyed the meaning(s) pleaded about her; and
3.such meaning(s) as proven are defamatory (this element is conceded if elements 1 and 2 are satisfied).
[70] Respondent’s closing submission, [187].
The law applicable to the present case differs from that considered by Blue J at first instance and by the Full Court on appeal, as detailed later in these reasons. For the purposes of this decision the question of liability is analysed in the manner proposed by Google, a course not opposed by Dr Duffy.
Publication
Legal principles
In this case as in Trkulja v Google LLC:[71]
In point of principle, the law as to publication is tolerably clear. It is the application of it to the particular facts of the case which tends to be difficult, especially in the relatively novel context of internet search engine results.
(Footnote omitted)
[71] (2018) 263 CLR 149, [39].
Defamatory matter is not published in the requisite sense until it is comprehended by a third party as explained by Kiefel CJ and Gleeson J in Google LLC v Defteros:[72]
Publication was described as a technical term, which is to be understood as a bilateral act by which the publisher makes the defamatory material available and a third party has it available for their comprehension. It may be understood as the process by which a defamatory statement or imputation is conveyed. Adopting Webb, any act of participation in the communication of defamatory matter to a third party is sufficient to make a defendant a publisher, regardless of their knowledge or intent. So understood, a person who has been instrumental in, or contributes to any extent to, the publication of defamatory matter is a publisher.
(Footnotes omitted)
[72] (2022) 403 ALR 434, [21].
As cryptically expressed by Cardozo CJ in Ostrowe v Lee,[73] publication:
is a term of art … A defamatory writing is not published if it is read by no one but the one defamed. Published, it is, however, as soon as read by anyone else.
[73] (1931) 175 NE 505, 505.
Every publication by reference to a different URL gives rise to a separate and independent cause of action.[74] Finally, the place where the tort of defamation is committed is ordinarily the place where the damage to reputation occurs:[75]
In the case of material on the World Wide Web, it is not available in comprehensible form until downloaded on to the computer of a person who has used a web browser to pull the material from the web server. It is where that person downloads the material that the damage to reputation may be done. Ordinarily then, that will be the place where the tort of defamation is committed.
[74] Brunswick, Duke of v Harmer (1849) 14 QB 185.
[75] Dow Jones & Co Inc v Gutnick (2002) 210 CLR 575, [44] (Gleeson CJ, McHugh, Gummow and Hayne JJ).
Hence a cause of action in defamation arises at the time when and in the place where the matter is identifiably comprehended, because “every communication of defamatory matter founds a separate cause of action” which is for the purposes of this case where a person accesses the material.[76] For practical purposes this means an applicant who fails to identify the individual time and place when a reader comprehends the defamatory material must prove underlying facts from which an inference of publication can properly be drawn.[77] Expressed slightly differently, an applicant can prove publication though facts from which it can be inferred that the words were brought to the attention of a third person when it is a matter of reasonable inference that the matter complained of was “actually seen and read by some third party”.[78]
[76] Dow Jones & Co Inc v Gutnick (2002) 210 CLR 575, [27], [44].
[77] Sands v Channel Seven Adelaide Pty Ltd (2009) 104 SASR 452, [391], citing Al Amoudi v Brisard [2007] 1 WLR 113, [32]-[37]. This decision was cited with approval in Capolingua v Nationwide News Pty Ltd [2016] WASC 156, [26]; Dods v McDonald (No 1) [2016] VSC 200, [9]; Stoltenberg v Bolton (2020) 380 ALR 145, [112].
[78] David v Abdishou [2012] NSWCA 109, [286], quoting P Milmo and WVH Rogers, Gatley on Libel and Slander (2008, Sweet & Maxwell, 11th ed).
The following propositions of relevance to this case were recently authoritatively laid down by the High Court in Google LLC v Defteros:[79]
·Internet search results including hyperlinks to defamatory material which are not themselves defamatory do not fall within the purview of publication because the hyperlink does not itself communicate anything and because following the link requires an act by a third party to access the hyperlink;[80]
·The fact of Google’s ability to remove a webpage identified by its URL from search results, that it operates for profit and whether it received notice are not relevant to the issue of whether its conduct amounted to publication for the purposes of defamation law;[81] and
·Notice of the existence of defamatory material may be relevant to knowledge in the defence of innocent dissemination.[82]
[79] (2022) 403 ALR 434.
[80] At [41]-[43], [53] (Kiefel CJ and Gleeson J), [59], [72]-[74] (Gageler J).
[81] At [12], [54] (Kiefel CJ and Gleeson J), [59], [62]-[64], [66] (Gageler J).
[82] At [55] (Kiefel CJ and Gleeson J).
Direct evidence
The direct evidence of publication led or given by Dr Duffy was quite thin and often inadmissible. The evidence such as it is, is only allowed to the extent that follows.
At first Dr Duffy called Mr Roberts purportedly as a person in South Australia who used a web browser to access Google sites, however she withdrew his evidence during the trial, so this came to nought.[83]
[83] T544.1-34.
During the interlocutory proceedings seeking permission to amend her statement of claim and to proceed with this action, Blue J received oral evidence from Michelle De Veau who undertook searches using Google in relation to Dr Duffy as relevant to proof of publication.[84] It was first proposed by Dr Duffy to use in this trial both Ms De Veau’s oral evidence before Blue J and her sworn affidavit tendered in the interlocutory process. Dr Duffy mistakenly assumed that these were automatically admitted in this trial. This assumption is, of course, not the case and it was opposed by Google. Before that could occur, it was necessary to seek admission under the business records provisions of the Evidence Act1929 (SA). None of those provisions apply to the circumstances. The same applies to Mr Barrow to whom Dr Duffy referred at times, but whom she did not call, so any secondary evidence of searches conducted by him is simply inadmissible.[85] The same considerations equally apply to the witness Mr Roberts with respect to the tender of his affidavit filed in the interlocutory proceedings concerning publication.[86]
[84] Duffy v Google LLC [2019] SASC 157, [17].
[85] T471.14-474.14.
[86] ATrB V6, p 1103.
References to the evidence of Ms De Veau previously given in the summary judgment hearing are, as explained above, simply inadmissible in this proceeding in the absence of calling her, or in the absence of proof that she was “unfit … to attend as a witness … and it is not reasonably practicable to secure [her] attendance”.[87] By the same token it is inappropriate to draw any inference from the failure to call her during this trial to the effect that she would not have assisted Dr Duffy’s case, given her intimation that she did not want to give evidence and that Dr Duffy was not directly in communication with her.[88]
[87] Evidence Act 1929 (SA) s 34C(1a).
[88] Closing Written Submission of Applicant, [60].
The relatively short evidence of Nadia Dries is more problematic. She deposed to searching Dr Duffy’s name using the Google search engine which produced a reference to a “psychic stalker” after clicking on a search result.[89] She did this just after learning Dr Duffy had won against Google in the first action.[90] That judgment, it is to be recalled, was handed down on 27 October 2015. Otherwise, her evidence was imprecise. It is open to infer that she saw one of the subject Ripoff Reports on google.com.au at some time shortly after the first judgment, but that is as far as it goes. It therefore amounts to but a small strand of circumstantial evidence which in and of itself does not demonstrate the probability of publication on a single occasion at relevant times.
[89] T601.20-29.
[90] T603.22-34, T605.2-23, T607.18-36, T608.4-16.
The evidence of Dr Duffy herself does not directly assist in proving publication except to the extent that she conducted the searches referred to above. Otherwise she was unwilling in the course of cross-examination to accept correspondence to the effect that in order to consider blocking URLs from google.com.au, Google required the specific URLs to be provided, or that Google would not block defamatory content under Australian law in results served on google.com.[91] She was inconsistent in her evidence as to the number of searches she herself conducted and as to the number of times she asked others to search for her.[92] The omission to take steps to pursue further removals was however borne of a growing sense of frustration and exasperation owing to Google’s insistence on nothing less than complete URLs before removals would be considered or made.[93] In any event this means that circumstantial proof of publication is entirely dependent on the evidence which Dr Duffy herself marshalled and as to the inferences that might be drawn therefrom.
[91] T416.37-417.37, T427.33-428.7, 428.26-32.
[92] T381.32-383.4, T465.38-467.25, T470.37-476.25, T644.34-646.8.
[93] T377.14-17, T359.24-27, T413.26-417.11, T418.4-5, T427.33-433.22 T438.4-10, T593.1-595.18, T661.27-663.14, T666.5-9, T667.36-668.8, T652.21-34.
There was no direct evidence whatsoever of publication on google.com.
Inferential evidence
google.com.au
Despite this conclusion as already noted, it nevertheless remains open to an applicant who fails to identify the individual time and place when a reader comprehends the defamatory material to prove instead publication by inference. In this context the court is “entitled to take into account, as a matter of judicial notice, that the use of the internet, to ascertain information about particular people, is commonplace”, so that the courts are not slow to draw an inference of publication on websites.[94] There is no rebuttable presumption to that effect.[95]
[94] Trkulja v Yahoo! Inc LLC [2012] VSC 88, [36]. See also Dods v McDonald (No 1) [2016] VSC 200, [9].
[95] Al Amoudi v Brisard [2007] 1 WLR 113, [37].
Documents provided by Google in the discovery/disclosure process in this action record its own data of all Google searches on google.com.au and google.com using the phrases “janice duffy”, “dr janice duffy” and “janice duffy reviews” between October 2015 and March 2018. Of those between October 2015 and October 2016, there were the following number of searches:[96]
·“dr janice duffy” – 4,535
·“janice duffy” – 4,576
·“janice duffy reviews” – 183
[96] Applicant’s Tender Book (ATeB) V2, pp 376-378.
Before embarking on an analysis of any inference that might duly arise from this data and the weight to be given to it, it must be acknowledged that it covers both google.com.au and google.com without differentiating the numbers attributable to each. Furthermore, the Google data does not disclose what the searcher did once the search results came up and it does not take into account that some users might have undertaken the same search more than once. On the other hand, it appears much more probable that the subject snippets would entice more Australian users to open the hyperlink because of the references to “Australian Psychic” and “Adelaide South Australia” for instance.
Some of these searches were obviously undertaken by Dr Duffy herself, as is known from the various annexures she produced and it is possible although unlikely there were more that she did not produce. It also seems likely that a number of searches were stimulated by her appearance on national television the day following judgment on “The Project” and perhaps by a number of media articles published after the judgment was delivered. The actual numbers month by month for the relevant period tabulate as follows:[97]
[97] ATeb V2, pp 376-377.
Month dr janice duffy janice duffy janice duffy reviews October 2015 2,042 1,704 29 November 2015 691 377 24 December 2015 723 552 26 January 2016 136 516 3 February 2016 169 208 7 March 2016 139 153 2 April 2016 42 134 1 May 2016 75 152 20 June 2016 89 117 43 July 2016 76 136 23 August 2016 62 157 0 September 2016 250 255 5 October 2016 41 115 0
As against all these considerations and excluding from the present analysis the figures for “janice duffy reviews”, it is clear that there were a significant number of searches, namely 3,746, during October 2015 for “janice duffy” and “dr janice duffy”. Although the number for September 2015 remains unknown, the number of searches falls to 1,068 in November 2015 and 1,275 in December 2015, before tailing off and fluctuating to varying degrees during 2016.
In light of the number of the above searches made within the limitation period of one year, coupled with Dr Duffy’s evidence of RoR1 appearing as the first ranking search result in all her searches and particularly the relatively large number of searches in the last quarter of 2015 on google.com.au, it is inconceivable that no user was shown the snippets or selected the underlying hyperlink to RoR1 in the period between 27 October 2015 and 6 November 2015. To the contrary, whilst it can be expected that some searchers might be more interested in those snippets reporting favourably as to Dr Duffy’s success in the first action, it is very probable that the majority were not simply interested in one aspect as opposed to both aspects of the case, especially so as to understand exactly what the Ripoff Reports were all about. Even in the case of those searches responsive to media publicity, RoR1 remained a highly ranked result, if not the first ranked search result and the user would more than likely be tempted to open the hyperlink to it.
As the Ripoff Report was clearly presented on the first search page in a preeminent position accompanied by snippets prompting, if not enticing, users to continue to the underlying Ripoff Report, it renders it more than likely that a not insignificant number of users conducting those searches in the relevant period clicked on the hyperlink and viewed RoR1. The only reasonable inference open is that many of those unidentified users did exactly just that, even though it is not possible to put a number on it except that the probabilities are in the scores, if not hundreds of such users. Of course, it would be less than that in the disclosure period between 27 October 2015 and 6 November 2015 but nonetheless a not insignificant number all the same. After all, the fact remains that searches for “dr janice duffy” and “janice duffy” peaked in that period, the majority of whom are more than likely to have followed the hyperlink to RoR1. On this process of reasoning, Dr Duffy has proven publication of RoR1 by inference.
Putting this analysis aside for the present, there is other material independently supporting the same conclusion. Google owns a function called “Keywords Planner”. This function allowed users to obtain data on the average number of monthly searches for words or combinations of words on a Google domain. It was so used by Dr Duffy to document retrospectively the 2015 data by means of a number of monthly searches for “janice duffy” and “dr janice duffy” which reveal that 6,500 searches were made relating to Dr Duffy in Australia in October 2015 and more than 1,000 searches in each of November and December 2015.[98] Here again the sheer number of such searches demonstrate it is highly likely that a significant number of searchers viewed RoR1. Expressed in another way, it is improbable in the extreme that not one of those searchers did not do so.
[98] ATrB V1, pp 214-216.
Google objected to the admission of the Keyword Planner material on the footing that the Google generated data referred to earlier is to be preferred because there is no proof that the data produced by the Keyword Planner is accurate and that the print outs are incomplete.[99] These criticisms may or may not be correct. This data is consistent with the conclusion of publication already reached. It is apparent that the material obtained by Dr Duffy using the Keyword Planner equally emanates from Google itself, so it was open to it to refute it if it wanted to. The defence called Dr Kondepudy, an expert software engineer with Google since September 2013, who said nothing about the provenance of the Keyword Planner data during his evidence, in his written statement[100] or in his written outline of evidence.[101] In that state of affairs the Keyword Planner material stands to be weighed according to the proof which was in the power of Dr Duffy to produce and in the power of Google to refute.[102]
[99] FDN 176, [11]-[14]; T125.19-23.
[100] Exhibit R7.
[101] Exhibit R7A.
[102] Blatch v Archer (1774) 98 ER 969, 970.
By parity of reasoning, whilst the 183 searches for “janice duffy reviews” between October 2015 and October 2016 are insufficient to draw the conclusion that scores of users accessed RoR2, it is impossible to conclude that none did. On this process of reasoning, Dr Duffy has likewise proven publication of RoR2 by inference, but in a relatively small number of cases.
google.com
The position is quite different when it comes to proof of publication in the case of the google.com searches. Dr Duffy provided evidence of 29 searches on google.com for “dr janice duffy” and “janice duffy” between 29 October 2015 and 19 October 2016, as set out in Annexure 4, which produced RoR1 as a search result. Dr Duffy also conducted eight searches on google.com for “janice duffy reviews” in the pleaded period, as set out in Annexure 8, which produced RoR2 as a search result. As was observed earlier, a user in Australia attempting to access google.com before October 2017 was redirected to google.com.au. It was of course possible for a more sophisticated user to override the default mechanisms, but in fact only a very small portion of users worldwide did so.[103] Nonetheless there is in any case no apparent reason and it is inherently improbable that a user in Australia would ordinarily bother to go to the trouble of bypassing the default regime erected by Google in order to directly access the US domain search engine.
[103] Exhibit R8, Tab 26 and RVK-1; T521.18-523.9, T688.14-690.6.
As reasoned earlier, there is no reason to suppose that anyone in the United States would have any particular interest in Janice Duffy of Adelaide, Australia. And the evidence of the minimal use of default mechanism overrides renders it unlikely that many if any users in Australia used the capacity to override.
All in all, the paucity of evidence fundamentally presents an insufficient substratum of facts to found an inference of publication in Australia through the medium of google.com. Dr Duffy’s claim for publication of RoR1 and RoR2 on google.com fails for want of satisfactory proof on the civil standard.
Capacity to convey defamatory imputations
Whether words or images complained of are capable of conveying a pleaded defamatory imputation is a question of law for the trial judge to determine.[104] As explained in the unanimous judgment in Trkulja v Google LLC:[105]
The test for whether a published matter is capable of being defamatory is what ordinary reasonable people would understand by the matter complained of. In making that assessment, it is necessary to bear in mind that ordinary men and women have different temperaments and outlooks, degrees of education and life experience. As Lord Reid observed in Lewis v Daily Telegraph Ltd, “[s]ome are unusually suspicious and some are unusually naive”. So also are some unusually well educated and sophisticated while others are deprived of the benefits of those advantages. The exercise is, therefore, one of attempting to envisage a mean or midpoint of temperaments and abilities and on that basis to decide the most damaging meaning that ordinary reasonable people at the midpoint could put on the impugned words or images considering the publication as a whole.
As the Court of Appeal of England and Wales observed in Berezovsky v Forbes Inc, that exercise is one in generosity not parsimony. The question is not what the allegedly defamatory words or images in fact say or depict but what a jury could reasonably think they convey to the ordinary reasonable person; and it is often a matter of first impression.
(Footnotes omitted)
[104] Trkulja v Google LLC (2018) 263 CLR 149, [30], citing Jones v Skelton [1963] 1 WLR 1362, 1370; Favell v Queensland Newspapers Pty Ltd (2005) 221 ALR 186, 189-190 (Gleeson CJ, McHugh, Gummow and Heydon JJ).
[105] (2018) 263 CLR 149, [31]-[32].
Whether a passage conveys a defamatory imputation is objectively assessed, that is, according to the understanding formed by ordinary, reasonable recipients of the communication as assessed by the trier of fact when there is no jury.[106] Still further as Kirby J observed in Chakravarti v Advertiser Newspapers Ltd, “[w]here words have been used which are imprecise, ambiguous or loose, a very wide latitude will be ascribed to the ordinary person to draw imputations adverse to the subject”.[107]
[106] Mirror Newspapers Ltd v Harrison (1982) 149 CLR 293, 298-301.
[107] (1998) 193 CLR 519, 574.
Once the judge determines that preliminary question, “the test of capacity of a published matter to defame is … whether any of the search results complained of are capable of conveying any of the defamatory imputations alleged”.[108] In defamation trials by jury, as was the case in Trkulja v Google LLC, this second question is a matter for the jury. In South Australia, juries were never empanelled in civil trials.[109] Justice Blue was not asked by Google in the first trial to consider both questions.[110] As observed in Trkulja v Google LLC,[111] Blue J “was describing the process of reasoning by which his Honour, sitting as trial judge, reached findings of mixed fact and law in the trial of a defamation proceeding before judge alone”.
[108] Trkulja v Google LLC (2018) 263 CLR 149, [52].
[109] Section 5 of the Juries Act 1927 (SA) provides ‘No civil trial is to be held before a jury’. This was historically the case even though the original Juries Act (No 1 of 1862) provided for juries of 12 men in civil case in s 26, a provision replicated in Juries Act 1917 (SA) s 134. A possible explanation for this may lie in the fact that qualification for jury service was limited to men between 21 and 65 who owned or occupied property of the yearly value of no less than £20: Juries Act 1917 (SA) s 5. This was lifted to £50 capital value in the 1927 legislation to be in line with the requirements to be eligible to vote: Juries Act 1927 (SA) s 11; South Australia, Parliamentary Debates, House of Assembly, 6 October 1927, p 941. According to Hague’s History of the Law in South Australia (University of Adelaide Barr Smith Press, 2005) “It can hardly be said that the jury system worked very satisfactorily. … In January 1839, out of a population of some 5,000, only 102 men could be found fit to be put on the list of special jurors, and 113 on the list of common jurors”: at p 554.
[110] T839.8-844.9, T878.2-880.9.
[111] (2018) 263 CLR 149, [36], referring to Duffy v Google Inc (2015) 125 SASR 437, [375] (Blue J).
Be that as it may, a conclusion that material is defamatory necessarily entails the prerequisite conclusion that the material holds the capacity to defame. It is illogical to suppose that a judge, finding a particular matter defamatory, could nevertheless conclude that matter was incapable of conveying a defamatory imputation.
Whether material is defamatory is judged by reference to “an ordinary reasonable person who has made the Google search in issue”.[112] The capacity of published matter to defame “must be assessed by reference to the most damaging meaning that could reasonably be put upon the words in question”.[113] And as Lord Halsbury LC noted in Lord William Nevill v The Fine Art and General Insurance Company, Limited, “it is necessary to take into consideration, not only the actual words used, but the context of the words, and the persons to whom the communications were made”.[114] Of course, it is essential that the subject search results be considered as a whole rather than individually, in isolation.
The snippets
[112] Trkulja v Google LLC (2018) 263 CLR 149, [53].
[113] Trkulja v Google LLC (2018) 263 CLR 149, [64], citing Lewis v Daily Telegraph [1964] AC 234, 259; Favell v Queensland Newspapers Pty Ltd (2005) 79 ALJR 1716, [17].
[114] [1897] AC 68, 72.
Defamatory meanings?
With these principles in mind, Google contends that the subject snippets taken in isolation are incapable of conveying any defamatory meaning. As mentioned earlier, the defamatory status of the Ripoff Reports is not in issue.
Returning to the first snippet:
Ripoff Report | Janice Duffy - Psychic Stalker! Psychics …
16, 2014 - Janice Duffy - Psychic Stalker! Psychics Beware Of Australian Psychic Stalker! Janice Duffy Stalked me on the computer for several monthsThe core matters in this are the generalisation “Beware Of Australian Psychic Stalker!” and the specific “Janice Duffy Stalked me on the computer for several months”. The noun “stalk” and its derivations carry an unmistakably pejorative connotation. The Oxford English Dictionary Draft Edition of 1997 in one entry describes a stalker as:
A person who pursues another, esp. as part of an investigation or with criminal intent; spec. one who follows or harasses someone (often a public figure) with whom he or she has become obsessed.
The Macquarie Dictionary ascribes the following meanings to “stalk”:
stalk
verb (i) 1. to pursue or approach game, etc., stealthily.
2.to walk with slow, stiff, or haughty strides.
3.to proceed with slow, implacable and often sinister movement: famine stalked through the land.
4.Obsolete to walk or go stealthily along.
–verb (t) 5. to pursue (game, a person, etc.) stealthily.
6.to harass (someone) by persistently and obsessively following them, telephoning them, etc.
7.to harass (someone) through the use of the internet, email, chat rooms, or other digital communications devices.
8.to investigate (someone) by conducting online searches, observing their social media posts, etc.
9.to stalk over or through.
–noun 10. an act or course of stalking game or the like.
11.a slow, stiff stride or gait.
–stalker, n
Stalking is also a criminal offence throughout Australia.[115]
[115] Criminal Code Act 1899 (Qld) s 359B(c)(ii); Criminal Code Act 1924 (Tas) s 192(1)(g)-(h); Criminal Law Consolidation Act 1935 (SA) s 19AA(1)(a)(iva)-(ivb); Crimes Act 1958 (Vic) s 21A(2)(ba); Crimes Act 1900 (ACT) s 35(2)(h).
It is no different for the second snippet:
Ripoff Report | Janice Duffy - Psychic Stalker! Psychics…
16, 2014 - *UPDATE Employee: Dr Janice Duffy Psychic Stalker Repatriation General Hospital Administrator Extorted. *UPDATE EX-employee responds: …and with the third snippet:
Ripoff Report | Dr. Janice Duffy Complaint Review
6, 2008 - Dr. Janice Duffy Complaint Review: Dr. Janice Duffy Stop the Australian Psychic Stalker Dr. Janice Duffy!! Adelaide, South Australia Adelaide,…The ordinary and natural meaning of these snippets clearly contains the capacity to convey defamatory meanings and in fact does so, namely that Dr Duffy covertly or stealthily, sinisterly and unlawfully stalked and harassed psychics and is someone to be fearful and beware of. It may be acknowledged that the “pages of search results contain a weight of other search results”[116] and yet the gravamen of each snippet which it must be recalled is highly ranked, is negative towards Dr Duffy. This is not a case in which refutation “destructive of the entire basis upon which the imputation relies” is present, for there is no such passage in the other snippets on the search pages containing the offending snippets.[117]
[116] Respondent’s closing submission, [248].
[117] Farquhar v Bottom [1980] 2 NSWLR 380, [33]-[34]. See also John Fairfax Publications Pty Ltd v Rivkin (2003) 201 ALR 77, [26]; Morosi v Broadcasting Station 2GB Pty Ltd [1980] 2 NSWLR 418, 419(6).
There is however a further consideration. In Duffy v Google Inc,[118] Blue J ascribed similar albeit not identical defamatory meanings, finding that the Reports conveyed (amongst other things) the defamatory imputations that Dr Duffy stalks and harasses psychics by persistently and obsessively pursuing them. In large measure these independently coincide with the above conclusions, although once again they are not entirely identical. The conclusions of Blue J were upheld or at least undisturbed by the Full Court in Google Inc v Duffy.[119] In those circumstances a single judge of this Court is bound by those conclusions despite the fact that the capacity to defame was not expressly considered – albeit inconsequently – as a preliminary issue.
[118] (2015) 125 SASR 437, [358]-[363], [413]-[423], [474].
[119] (2017) 129 SASR 304, [80], [83], [383]-[387], [562].
Enticement or encouragement
The present case is distinguishable from the High Court decision in Google LLC v Defteros,[120] in as much as the latter was concerned with snippets that did not contain defamatory material. The hyperlink in question was considered “content‑neutral”.[121] The snippet in question in Defteros was this:[122]
Underworld loses valued friend at court‐SpecialsGanglandKillings ….
> Features > Crime & Corruption ▾June 18 2004 — Pub bouncer‐turned‐criminal lawyer George Defteros always prided himself on being able to avoid a king hit — The Age Online
[120] (2022) 403 ALR 434.
[121] Google LLC v Defteros (2022) 403 ALR 434, [43], citing Crookes v Newton [2011] 3 SCR 269, [30].
[122] Reproduced at [13].
In Defteros it was authoritatively determined that a hyperlink which is not itself defamatory does not fall within the purview of publication because that hyperlink did not communicate anything, even if it links to defamatory material.[123] This conclusion is founded on an insufficient connection between Google displaying search results in a user’s browser and the third-party website, which is hosted on that third-party URL and created by that third party.[124] Thus, Google had merely “assisted” persons searching the Web to find and access information.[125] It did not direct the user to a particular result and it was entirely a decision for the searcher whether to take the further step of clicking on the hyperlink so as to access the defamatory webpage.[126]
[123] Google LLC v Defteros (2022) 403 ALR 434, [41]-[43], [53] (Kiefel CJ and Gleeson J), [59], [72]-[74] (Gageler J).
[124] Google LLC v Defteros (2022) 403 ALR 434, [47]-[50], [53] (Kiefel CJ and Gleeson J), [72] (Gageler J).
[125] Google LLC v Defteros (2022) 403 ALR 434, [49] (Kiefel CJ and Gleeson J), [219] (Edelman and Steward JJ).
[126] Google LLC v Defteros (2022) 403 ALR 434, [49], [50] (Kiefel CJ and Gleeson J), [74] (Gageler J), [221] (Edelman and Steward JJ).
This difference was specifically acknowledged by Gageler J in Defteros:[127]
Thus, in Google Inc v Duffy with reference to Hird v Wood, the content of the snippet component of several Google search results was said to have been such as to “entice” a third party to click on the hyperlink in those search results to obtain more information. That was because the snippet component “naturally invite[d] the reader to click on the hyperlink for explanation and elaboration” of the matter, such as to be aptly described as “the electronic analogue of the person who places a post‐it note on a book which reads ‘go to page 56 to read interesting gossip about X’”.
(Footnotes omitted)
Gageler J further added:[128]
Unlike the position in Duffy, no feature of the content of the particular organic search result in the present case has been found to have operated as an enticement or encouragement to click on the hyperlink
[127] At [67].
[128] At [71]. See also Edelman and Steward JJ, [217].
A further consequence of Defteros may be whether search engine publishers such as Google can any longer be characterised as a secondary or subordinate publishers as a matter of law. Hitherto intermediate appellate courts have accepted that a search engine displaying results on a user’s browser is a secondary publication and does not therefore constitute a primary publication of those results.[129]
[129] Google Inc v Duffy (2017) 129 SASR 304, [181]-[184], [597]-[598]; Google Inc v Trkulja (2016) 342 ALR 504, [349], [353], [357].
Even so, Kiefel CJ and Gleeson J acknowledged that “It is of course possible that search results may themselves contain matter which is defamatory”.[130] Their Honours added:[131]
[T]he rules of publication apply where a person endorses, adopts or otherwise approves of defamatory matter which is to be published.
On that topic Gageler J wrote:[132]
To accept that the provision of a hyperlink is not enough to amount to participation in the process of publication which is completed when a third party clicks on the hyperlink so as to view the webpage, however, is not to deny that the provision of a hyperlink might combine with other factors to amount to participation in that process of publication of matter on that other webpage … The question whether particular action amounts to enticement or encouragement of that nature is appropriately described as one of “fact and degree”.
(Footnotes omitted)
[130] Google LLC v Defteros (2022) 403 ALR 434, [23].
[131] Google LLC v Defteros (2022) 403 ALR 434, [45].
[132] At [66].
For their part Edelman and Steward JJ characterised the snippet at issue in Defteros in the following way:[133]
It may be accepted that, if a search result, by its content, was likely to entice a searcher to choose a given third‐party webpage over others, then an inference might be drawn that the assistance in the search was given with a common intention shared with the third party in the sense of a concerted action to a common end. However, here, even that evidentiary foundation was absent.
Based on the foregoing finding, the words that accompanied the search result simply described a story about a criminal lawyer and his clients, and no more. They indicated what the Underworld article would contain. Of course, for the reasons already given, it may be accepted that the very purpose of the search engine is that it will produce search results that might be accessed by a given user. Depending on what a user is searching for, and why, words that indicate the contents of a given webpage listed in the search results may encourage that person to click on the hyperlink. But this is very far from a basis for a finding of enticement.
On the other hand, as to the intention to publish, they considered:[134]
It is an intention that the act of communication be performed or an intention to authorise, assist, or ratify the act of communication being performed by another; it is not concerned with the consequences of the act or with issues of fault, such as the reasonableness of the act.
(Footnote omitted)
[133] At [229]-[230].
[134] Google LLC v Defteros (2022) 403 ALR 434, [202].
The evidence was that Google further removed two URLs around 18 December 2017 after receiving a second statement of claim on 14 December 2017 containing the complete URLs,[179] another one around 7 March 2018,[180] and eight URLs on 8 or 9 July 2018, at least six of which were included in an affidavit of Dr Duffy filed on 2 July 2018.[181] Google’s position is that once it acquired knowledge of the defamatory URLs and then failed to remove them within a reasonable time, the required mental element for the defence cannot be established because the “continuing existence thereafter was the direct result of human action or inaction rather than merely the result of machine operation”.[182] What was reasonable for this purpose must be considered in the context of the past conduct of Google in completely failing to remove offending URLs before the first action was instituted by Dr Duffy in February 2011, by “making Dr Duffy jump through hoops, and then back through the same hoops” thus making “the process for having search results removed so difficult as to deter her from persisting”.[183]
[179] Exhibit R8 V2, pp 806, 810; Exhibit R7B, items 50, 54.
[180] Exhibit R7B, item 49.
[181] Exhibit R7B, items 42, 43, 44, 45, 47, 51, 52, 53. Dates sometimes vary by a day or so because of time differences between the US and Australia.
[182] Google Inc v Duffy (2015) 125 SASR 437, [206].
[183] Google Inc v Duffy (2017) 129 SASR 304, [484]. See T417.9-34, T428.3-430.15, T500.1-16, T593.1‑594.1.
The persistent stance emerging from the pattern of behaviour by Google was that it considered “knowledge” for the purposes of innocent dissemination to equate with precise knowledge of the offending URL, rather than knowledge that a particular matter was defamatory. In reality, the position was neither as simple nor as straightforward as that. Section 14(2) of the Defamation Act as it stood at relevant times merely required a concerns notice to be in writing. By s 14(3) the onus was on Google as the publisher to “give the aggrieved person a written notice … requesting the aggrieved person to provide reasonable further particulars about the imputations of concern as specified in the further particulars notice”. Reasonable further particulars and complete URLs are not the same requirement. This is in stark contrast to the present-day s 12A(1)(a)(ii) which whilst falling short of requiring the provision of complete URLs in the concerns notice requires that a concerns notice “specifies the location where the matter in question can be accessed (for example, a webpage address)”.
The pattern of removals adopted by Google in relation to Dr Duffy were as follows. It begins with the first four removals on 7 March 2011 as follows:[184]
[29] Exhibit R7B, items 29, 30, 31, 33.
Next are the 16 removals of 31 May 2011, as follows:[185]
[185] Exhibit R7B, items 1, 2, 10, 14, 15, 21, 23, 26, 27, 32, 34, 35, 38, 39, 41, 60.
[1]>
The third tranche of 13 removals was around 2 November 2011, as follows:[186]
[186] Exhibit R7B, items 4, 5, 6, 7, 8, 11, 12, 13, 16, 19, 20, 22, 37.
[4]>
These were followed by 15 further removals on 14 November 2011, as follows:[187]
[187] Exhibit R7B, items 9, 17, 18, 24, 25, 48, 55, 56, 57, 58, 59, 61, 62, 63, 64.
[9]>
Finally, a further three removals were affected before 27 October 2015 on 29 November 2011:[188]
[3] Exhibit R7B, items 3, 36, 40.
Further removals were affected following the delivery of the liability judgment: two on 5 November 2015, two on 18 December 2017, one on 7 March 2018, and the remaining eight removals on either 8 or 9 July 2018.[189] However all but the November 2015 removals lie outside the limitation period ending on 27 October 2016. There is also one undated removal.[190]
[189] Exhibit R7B, items 46, 65, 50, 54, 49, 42, 43, 44, 45, 47, 51, 52, 53 respectively.
[190] Exhibit R7B, item 28.
Each of these 53 acts of removal between 7 March 2011 and 5 November 2015 served as effective renotification to Google of the continuing publication of material defamatory of Dr Duffy in line with those identified in the Notice of Concerns of 28 April 2014.[191] Most of the URLs involved in those removals contained specific reference to Dr Duffy and of course were supplied to Google by her or her lawyers. Each one was a reminder that the offending material relating to Ripoff was re-indexed on Google with slight changes to the URL. Each one constituted in effect constructive renotification to Google as a secondary publisher of both the search results and the underlying Ripoff webpages. It can be seen in many cases the URL changes or iterations over time made by Ripoff were quite small. For the most part, the high degree of similarity or congruence in the characters contained in the removed URLs serves to confirm the accuracy of the opinion expressed by Mr Silver-Smith that “elided URLs … provided … sufficient information to rapidly and easily identify the pages in question” and that “Google can effect removals without requiring that all URLs of contravened content be provided to them”. All Google need to effect removal was the search term “janice duffy”.[192]
[191] ATeB V1, pp 1-32.
[192] T711.35-712.13.
There is no doubting that Google knew of the existence of the defamatory material for the purposes of this case by no later than 27 October 2015 and yet it did nothing to search, locate or block the URLs when it should have. In any case on the basis of the proven facts, it plainly ought reasonably to have known that the reassigned Ripoff Reports contained defamatory material. It was clearly negligent in failing to take any steps of its own accord to ascertain the subject URLs other than to insist that Dr Duffy identify them fully for Google, knowing that in the meantime the offending material remained accessible and was destined to reappear in a short space of time.
It is very apparent that Google moved to block or remove offending URLs only once given complete or unelided URL sequences. By doggedly and unrealistically insisting on nothing less despite having the means at hand to easily locate them if it had wanted to, it acted unreasonably. It was entirely reactive rather than proactive in the removals process. In truth Dr Duffy and her legal advisors (when they were engaged) were effectively stuck on a never-ending treadmill from which she could not escape of identifying complete URLs, securing removal by Google only to find the same posts with altered URLs inexorably reappearing whilst Google stood by doing nothing itself. It is to be recalled in this context that the evidence of Mr Silver-Smith was that “Google could have fairly easily created a method to detect Ripoff Report’s subversion, and automatically counteracted it”, evidence which Google at no point countered or disproved.[193]
[193] Exhibit A1A, [112].
Therefore, the fact that only an elided URL for RoR1 and no URL for RoR2 was provided to Google does not mean that Google did not have the requisite knowledge. In the above circumstances Google singularly fails to prove that it did not or that it could not reasonably have known the matter was defamatory and that its lack of knowledge was not due to any negligence on its part.
Google as an innocent disseminator?
For the reasons advanced above it must follow that the defence of innocent dissemination is unavailable to Google.
Triviality defence
Section 31 of the Defamation Act provided:
31—Defence of triviality
It is a defence to the publication of defamatory matter if the defendant proves that the circumstances of publication were such that the plaintiff was unlikely to sustain any harm.
By the Defamation (Miscellaneous) Amendment Act 2020 (SA) commencing on 1 July 2021, s 31 was repealed consequential to the inclusion of s 10A which added a serious harm element to the defamation cause of action. However, as Part 3 of Schedule 1 provides, those amendments to the Defamation Act apply only in relation to the publication of defamatory matter after the commencement of the amendment. Therefore, no question of retrospectivity arises and the defence remains available to Google.
This defence requires retrospective consideration of the circumstances existing at the time of publication, that is, it requires an inquiry as to the likelihood of harm, concerned as it is with the “the quality of the publication in respect of its proneness to cause harm”.[194]
[194] Barrow v Bolt [2015] VSCA 107, [34], quoting Chappell v Mirror Newspapers Ltd (1984) Aust Torts Rep 80-691, 68-947.
The remaining principles surrounding the triviality defence were summarised by Kaye JA:[195]
Secondly, the focus of the inquiry is on the ‘circumstances of the publication’. The critical test is whether those circumstances were such, at the time of publication, that it was unlikely that the applicant would suffer harm. The circumstances include (inter alia) the content of the publication, the extent of the publication, the nature of the recipients and their relationship with the applicant. However, the phrase ‘circumstances of the publication’ is not sufficiently wide to encompass the previous bad reputation of a plaintiff.
Thirdly, the phrase ‘unlikely to sustain any harm’ does not mean that it is sufficient for the defendant to establish that it is ‘more probable than not’ that the plaintiff will not suffer harm. Rather, the defendant must demonstrate that there is ‘the absence of a real chance’, or the ‘absence of a real possibility’, of harm.
Fourthly, the defendant is required to establish that, at the time of publication, the circumstances were such that the plaintiff was unlikely to suffer ‘any’ harm. Accordingly, the onus, on the defendant, to prove that matter, is high.
Fifthly, the defence, provided by s 33, applies to the publication of ‘defamatory matter’. Thus, s 33 provides a defence where matter, that has been published, is defamatory of the plaintiff. In order to be defamatory, the matter must tend to lower the reputation of the plaintiff in the eyes of ordinary reasonable members of the community. Thus, s 33 contemplates a case in which, notwithstanding that a publication about a plaintiff is defamatory in that sense, nevertheless the ‘circumstances of publication’ were such that the plaintiff was unlikely to sustain any harm as a result.
(Footnotes omitted)
[195] Barrow v Bolt [2015] VSCA 107, [35]-[38], referring to Morosi v Mirror Newspapers Ltd [1977] 2 NSWLR 749, Jones v Sutton (2004) 61 NSWLR 614.
Google’s stance is that any publication was unlikely to cause any harm to Dr Duffy in the context of all relevant circumstances, including:
·That her reputation was vindicated by success in the first action in respect of the sole imputation upon which she sues in this case;
·The favourable publicity surrounding the first action, including newspaper articles in 2011 and 2012,[196] in the media and in the legal and academic professions,[197] that she in fact spoke to media outside Court following delivery of the judgment on 27 October 2015 and she was interviewed on the national television show ‘The Project’ the following night;[198]
·That it can be inferred from the pattern of searches coinciding with media publicity surrounding the liability judgment in the first action that the majority cause of such searches between October and December 2015 were performed most likely by persons necessarily having some knowledge of Dr Duffy, as they will have done so knowing that a Court found similar content to be false and defamatory of Dr Duffy; and
·That RoR1 was blocked on google.com.au from 5 November 2015 onwards so that no user could have viewed RoR1, absent the prospect of any such user overriding Google’s default to access its US country service at google.com.
[196] Exhibit R9.
[197] Exhibit R8 V3, tabs 23, 24; T367.30-33.
[198] T459.13-17; Exhibit R8 V3, tab 21; T367.34-38 respectively.
Applying the above principles to the proven facts, the content of the snippets referred to were defamatory in the sense explored earlier and the hyperlinked Ripoff Reports were indeed pernicious. For RoR1, the extent of publication was extensive in terms of the number of readers, as disclosed by Google’s own data of Google searches (supported by the Google Keyword Tool data) and the snippets in question were highly or prominently ranked in the search results. Here again the Google Keyword Tool data showed 6,500 searches were made relating to Dr Duffy in Australia in October 2015 and approximately 1,000 searches in November 2015 which affirms that dissemination was significant.[199] Of course, the extent of publication between 27 October and 5 November 2015 would necessarily be restricted by the short time frame, but that must be viewed in the context of an intense period of searches leading to RoR1. While the publication of RoR2 was much less extensive, with only 183 searches in the pleaded period, RoR2 was still highly ranked and it was available on google.com.au for a significant amount of time. As it is unknown who the recipients were, any consideration of their relationship to Dr Duffy is not possible or is at best entirely speculative and even if there were searchers who had “some knowledge of Dr Duffy” it does not follow that they were necessarily supporters of her, and each such search nevertheless is an occasion of publication.
[199] ATrB V1, pp 214-216.
Dr Duffy’s email to Professor Hennessy of 3 October 2016 does not simply vent her enormous frustration with Google, it reveals the contemporary consequences to her as a result of the subject matter of these proceedings, in that “my life and career [are] destroyed”, and as an expression of her situation at the time:[200]
I was tired, emotional and I was basically saying ‘Well, I am going to have to sue again if it's not removed’. It was a last ditch attempt to get it removed before I moved on to filing proceedings … because they clearly have tried to punish me for litigating’ which I felt I still do …
[200] ATeB V2, p 349; T257.13-21.
Questions of vindication and benefitting from substantial favourable publicity serve only to do less than justice to an aggrieved party in awards of damages for defamation. The House of Lords accepted this proposition in Associated Newspapers Ltd v Dingle:[201]
Such a method of assessing damages would do less than justice to the plaintiff, in my view, and it is based upon suppositions which may be unfounded. A judge cannot tell how widely his judgment will be reported and read, nor can he tell how far the plaintiff’s general reputation will be improved by his complimentary remarks. A simple verdict of a jury in favour of the plaintiff will no doubt have a good effect on his reputation, and it is surely impossible to set a monetary value upon the difference, if any, between the effect of a jury’s finding and the effect of a judge’s finding plus a compliment from him.
[201] [1964] AC 371, 404 (concurrences at 400-401, 407, 408-409, 419).
Applying this decision, White J reasoned in Hockey v Fairfax Media Publications Pty Ltd:[202]
First, damages are the principal means by which the Court speaks to provide vindication. Secondly, although the outcome of the cases may well attract considerable publicity, it may be doubted that members of the public generally will read these reasons. It is more likely that they will have regard to the “headline judgment” constituted by the awards. Thirdly, it would be undesirable to introduce, in effect, two tiers of damages for defamation: the first when the case is tried by a judge alone and reasons published, and the other when the case is tried by jury and no reasons published.
[202] (2015) 237 FCR 33, [501].
Given the fractious and if not at times hostile relationship between Dr Duffy and Google, as well as Dr Duffy’s adverse reaction to the publications found to be defamatory in the first action, it was inevitable that she would experience a significant adverse reaction to the continued publication of that material under a different URL. It must follow that Google fails to demonstrate that there is “the absence of a real chance” or alternatively the “absence of a real possibility” of harm to Dr Duffy.
Furthermore, it was at the time of publication foreseeable that absent removal, Dr Duffy was likely to relitigate, a course that clearly materialised. In that situation the prospect of hurt and distress suffered occasioned by litigation and trial is compensable as forming a portion of the harm caused by defamatory publications.[203] The findings and acceptance of the evidence of Dr Duffy (with some exceptions) by Blue J and the level of damages he awarded is some measure of the foreseeable adverse reaction likely to be suffered by Dr Duffy in the event of republication.[204]
Dr Duffy gave evidence of the effect of the publication by Google of the defamatory material upon her. She said that she felt absolutely sick, mortified and humiliated when she became aware of the defamatory material and those feelings continued throughout 2010 and to the present day. She suffered extreme embarrassment and distress knowing that the defamatory material was on Google’s webpages and would be seen by anyone searching her name using Google, including personal and professional acquaintances. Her distress became worse after she notified Google in September 2009 and Google failed to remove it, as her situation felt hopeless. I accept this evidence, which was not in any event challenged in cross-examination.
[203] Humphries v TWT Ltd (1993) 120 ALR 693, 706.
[204] Duffy v Google Inc (No 2) [2015] SASC 206, [109].
Miscellaneous
At times Dr Duffy was prone to direct criticism of Google for taking the inconsistent stance that in the United States it functions in a capacity similar to that of a newspaper or a literary publisher,[205] whereas in Europe, Australia and other common law countries it contends the automated nature of its search algorithms relegates it to the position of a passive distributor and therefore cannot be liable for defamatory publication on its domains even after notification.[206]
[205] Citing Search King v Google Technology Inc., 2003 WL 21464568 (W.D. Okla. 2003).
[206] Metropolitan International of Schools Ltd t/as Skillstrain and/or Train2Game v Designtechnica Corp t/as Digital Trends [2011] 1 WLR 1743 (QB).
So far as this contention is concerned, it should be understood that with both search engines and country domains before and after October 2017, Google takes the law as it finds it in the jurisdictions in which it operates.
Dr Duffy was equally critical of Google’s selective use of United States law as a justification for failing to remove defamatory content and its stance that it was not required to remove defamatory material from google.com despite it being accessible in Australia, under the guise of freedom of speech protections. She evidenced the removal in 2013 of defamatory hyperlinks and snippets posted on RipoffReport.com about a man who lived and worked in Melbourne from both google.com and google.com.au.[207] It transpires that this came about under valid US court orders for removal, so this provides no basis for the proposition that Australian courts have the power to order the removal of webpages on google.com, absent the presence of jurisdiction accruing from defamatory material viewed in Australia.[208] Nor does it present any evidence of the capacity of Google to remove material, for it is accepted that it can. The question is whether it must by reason of the operation of domestic law in the jurisdiction in which defamatory material is published.
[207] Closing Written Submission of Applicant, [55]-[61].
[208] T435.6-436.13; Respondent’s closing submission, [117].
Conclusion
In summary, both RoR1 and RoR2, as well as all three snippets on google.com.au were defamatory and were published by inference to unknown third parties. There was no sufficient evidence that RoR1 or RoR2 were accessed on google.com in Australia and hence it cannot be accepted on the balance of probabilities that they were published in Australia. Google is proven to have participated in the communication of both RoR1 and RoR2 in Australia on google.com.au so as to render it liable as a secondary publisher. The defences of innocent dissemination and triviality fail. The parties are to be heard on the issue of consequential damages, before questions of costs arise.
0
25
1