Google and Online Liability

December 19, 2012

The issue of service provider web site liability continues to be a contentious area. There are various cases in the UK and the Court of Appeal judgment in Tamiz v Google is not far off.  

Notwithstanding the three defences in the eCommerce Directive, Mr Justice Eady went further. In Tamiz v Google [2012] EWHC 449 (QB), it was decided that an unrepresented claimant could not proceed to discovery and trial on the basis that the web site service provider was not a publisher at common law. This goes further than even a service provider could have hoped. If the original decision is correct, one wonders why there is a need for the three defences in the eCommerce Directive at all. 

The ground has also shifted since the original hearing before Eady J. Mr Tamiz obtained representation. There has been a public and political sea-change in perceptions in relation to the media and the Internet (eg hacking, trolls, invasions of Royal privacy, abuse, suicides, etc.). Case law has also advanced, especially in other jurisdictions. It is instructive to look at the approaches elsewhere to similar and related issues. 


In Milorad Trkulja v Google Inc and Google Australia, No. 10096 of 2009, [2012] VSC 533 (12 November 2012, Beach J) Google was found to be a publisher in an Australian defamation case. It was subsequently ordered to pay the defamed plaintiff AU$200,000. The plaintiff was being linked with false allegations and imputations of criminality and gangland activity, despite having objected and reported this to Google. 

The judgment summarises the arguments: 

‘While much was made by counsel for Google Inc of the fact that there was no human intervention between the request made to the search engine and the publication of search results, and of the fact that the system was “fully automated”, the plaintiff’s point was that Google Inc intended to publish everything Google’s automated systems (which systems its employees created and allowed to operate) produced. Specifically, the plaintiff contended that Google Inc intended to publish the material complained of because while the systems were automated, those systems were the consequence of computer programs, written by human beings, which programs were doing exactly what Google Inc and its employees intended and required. On this basis, it was contended that each time the material complained of was downloaded and comprehended, there was a publication by Google Inc (the operator and owner of the relevant search engines), as intended by it. So it was submitted by the plaintiff that Google Inc was a publisher throughout the period in respect of which complaint was made’ (at [16]).

 Beach J states (at [18] and [20]): 

‘The question of whether or not Google Inc was a publisher is a matter of mixed fact and law. In my view, it was open to the jury to find the facts in this proceeding in such a way as to entitle the jury to conclude that Google Inc was a publisher even before it had any notice from anybody acting on behalf of the plaintiff. The jury were entitled to conclude that Google Inc intended to publish the material that its automated systems produced, because that was what they were designed to do upon a search request being typed into one of Google Inc’s search products. [emphasis added] …. 

As was pointed out by counsel for the plaintiff in his address to the jury, the first page of the images matter (containing the photographs …) was a page not published by any person other than Google Inc. It was a page of Google Inc’s creation – put together as a result of the Google Inc search engine working as it was intended to work by those who wrote the relevant computer programs. It was a cut and paste creation (if somewhat more sophisticated than one involving cutting word or phrases from a newspaper and gluing them onto a piece of paper). If Google Inc’s submission was to be accepted then, while this page might on one view be the natural and probable consequence of the material published on the source page from which it is derived, there would be no actual original publisher of this page.’ 

Referring to the High Court’s judgment in Tamiz, Beach J states (at [27] and [28]): 

‘Eady J … does not appear to have given any consideration to the fact that internet search engines, while [sometimes] operating in an automated fashion from the moment a request is typed into them, operate precisely as intended by those who own them and who provide their services …. 

To say as a general principle that if an entity’s role is a passive one then it cannot be a publisher, would cut across principles which have formed the basis for liability in the newsagent/library type cases and also in those cases where someone with power to remove a defamatory publication chooses not to do so in circumstances where an inference of consent can be drawn’ (emphasis added). 

The judge also said that it was ‘doubtful that that same description [of internet intermediary] can be applied to an internet search engine provider.’ Beach J said that ‘such ‘internet intermediary’ is, in any event, performing more than the ‘merely passive role … [of] facilitating postings’ (at [29]). He again states that Google ‘was a publisher – even if it did not have notice of the content’  (at [30] and [31]). 

‘The question is whether, after relevant notice, the failure of an entity with the power to stop publication and which fails to stop publication after a reasonable time, is capable of … publication. Such an inference is clearly capable … Google Inc’s submission on this issue must be rejected for a number of reasons, the least of which is that it understates the ways in which a person may be held liable as a publisher.’  

Beach J said (at [37]) that, notwithstanding the Google tactical protestations, Google Inc was well aware of what was being requested of it. The judge also notes (at [35]): 

‘Google Inc’s witness … conceded … (perhaps somewhat begrudgingly) that it would not take very much effort to work out … [what] to do … it would have been open to Google Inc to block the URL of that page from Google Inc’s searches, in compliance with the plaintiff’s … request.’ 

In Trkulja v Yahoo! Inc & Anor [2012] VSC 88 the same plaintiff was awarded damages of $225,000 in respect of the publication of similar articles. The web site service provider was a publisher. 

New Zealand

In the New Zealand High Court, Abbot J in A v Google New Zealand [2012] NZHC 2352, another defamation case, also said that Google could be a publisher. Abbot J stated: 

‘it is reasonably arguable that a search engine is a publisher in respect of specific URLs and words …. The application for strike out is dismissed on the grounds that this is a developing area of the law and that it is arguable that the defendants are publishers.’


In the Criminal case of Drummond, De Los Reyes, Fleisher and Arvind, Milan, Docket No 14667/08, Judgment No. 1972/2010, there was a criminal prosecution for privacy and data protection breaches when an abusive video of a disabled student was published and not taken down, even after complaints. The case (now subject to a current appeal) is notable as Google executives were actually arrested.  

The Italian Google case is arguably misunderstood. This article is the first to examine an actual translation of the case. The decision deserves greater attention than the superficial commentary to date. The case judgment comprises a substantial 108 pages.  

One of the more interesting parts of the judgment arises from the Irish connection. One of the reasons why the Italian Court made its decision was because of evidence heard in the Dublin Metropolitan District Court as part of a rogatory request from Italy. Various witnesses were called before the Dublin Metropolitan District Court and questioned. The transcript was then sent to Italy. The evidence in Dublin confirmed control and relevant activity in Dublin as regards the offending video. At least four witnesses were heard at a hearing in Dublin before Judge Reilly on 28 May 2008. 

One reply (at p 33) indicated that:

‘We have created new tools to make it easier to flag the contents … that could be inappropriate. These tools also allow the creators of the contents to automatically remove material that breaches their own copyright. In addition, we also carry out a limited pre-screening. When we realize that any content breaches these guidelines – or when someone brings it to our attention – we remove it as soon as possible.’ 

The suggestion that ‘Google is taking the easiest path, asking users to “do the dirty work” and to flag contents under copyright or obscene contents’ was denied, with the Google witness saying (at p. 35):

‘we carry out a limited pre-screening and we have tools to check the videos that have a very high number of playbacks (as often happens with pornography). If we realize that the content breaches our guidelines we remove the content as soon as possible. We also encourage users to notify us when they find illegal contents’ 

However, there is an Italian Court footnote after the word ‘playbacks’ above, stating that ‘This … does not correspond with the truth, as emerged from the facts of these proceedings.’ Other Google witness statements are also criticised as ‘not correspond[ing] to the truth.’  

Video monitoring was found to have occurred in Dublin by Dublin employees. There was also a meeting in Dublin relating to the Italian abuse video (and another Italian abuse video). 

At one stage Judge Reilly had to say: 

‘perhaps I should remind the witness that he has sworn to tell the truth, the whole truth and nothing but the truth. I am hearing many words from you. You are here, I believe, in order to give us proof and information. I am just reminding you that from where I am sitting I am only hearing clichés, generic statements‘ (at p. 50, highlight added by Italian court). 

It appears the Public Accounts Committee is not the first to hear rehearsed statements. 


In Peterson v Google (03.09.2010, ref. no. 308 O 27/09), Google and YouTube were held liable for copyright infringement. There is also a further Google case currently being litigated in Germany (Mosley). 

YouTube (now a Google company) has also been sued by a German copyright collection society, GEMA. In GEMA v YouTube (20 April 2012, File reference 310 O EUR 461/10) the court held that there could be liability for infringement, particularly after notice was received, stating: 

‘through the provision and operation of the video [web site], they have contributed to the infringement. Because of this contribution, the defendants have conduct- and control- obligations. These have not been observed … and therefore [YouTube is] in breach of [its] duty of care.

Thus, the defendant was … in breach of [its] duty, to disable the relevant video clips without delay, after [it] had been informed by the claimant of the copyright infringement. With respect to the seven videos in question; disabling was only undertaken a good one and a half months after the notification was made by the claimant. In such period of time, it can no longer be referred to as an action without delay.’  

Referring to tools and technologies existing and available to Google and YouTube, the court states: 

‘it would be expected, that upon receipt of notification of copyright infringement, that future uploads, which matched the notified musical recording, would be prevented through the use of software. A software program, appropriate for this purpose, was available to the Defendant in the form of a Content-ID program which they had developed. The defendant must however apply the said program themselves, and not, as they contend, delegate this to their rights’ holder.’ 

Other Storm Clouds

One of the frequent arguments of certain service providers, including Google, is that abusive and infringing content occurs automatically via algorithms and third parties. However, the Federal Court of Australia (Australian Competition and Consumer Commission v Google Inc [2012] FCAFC 49, 3 April 2012) emphatically disagreed, holding against Google in relation to controlling and manipulating results. The court concluded that it was Google’s conduct as a principal that was misleading and stated that rankings were determined by Google.

 Google is also under pressure from potentially detailed contentious competition investigations in both the EU and US, partly on foot of complaints in relation to manipulating results to the detriment of competitors. 

There is a massaged impression from web sites that there is a global ecommerce defence applicable, regardless of the activities. Another massaged suggestion is that the defence(s) are not limited defences but apply generally, regardless of activity, jurisdiction, commerciality, profit, advertising, web site role, involvement, manipulation, etc., and that the defences apply without qualification, limitations, or possibility of being lost. 

The three eCommerce defences are, however, limited. It is up to a web site service provider to establish that its specific activity, or one of its activities, can fall within a given defence.  Each defence is different. Each is specifically designed and targeted at a specific activity.  It is possible for a specific activity of a web site to fall under one of the defences, but then to fall out of or to lose that defence. Any given web site activity cannot fit within more than one of the defences. These issues are complex, but are arguably becoming more complex and nuanced as more of these cases come to be litigated. A considered and detailed analysis is needed in each instance, not a massaged interpretation, even if verbally seductive. 

The Spanish ECJ case (Google Spain, S.L., Google Inc. v Agencia Española de Protección de Datos, Mario Costeja González, (Case C-131/12) and the Tamiz Court of Appeal judgment could also assist in interpreting liability and applicability of defences. Albeit, strictly speaking, the Tamiz appeal is really about the right to have one’s normal trial day in court.   However, it is also about victims of abuse, from defamation and privacy breach to trolling, and their ability to have their day in court when they need to. The wider unforeseen policy effect of the decision of Eady J would be to deny all victims, and indeed others. Was it an error? Was it overbroad? Australia, New Zealand, Germany, Italy, amongst others, and even a recent Belfast Facebook case (XY v Facebook [2012] NIQB 96 McCloskey J), hold that the service provider web site can be a publisher. The success of the Tamiz appeal would merely establish, or re-establish, that all of these cases remain to be argued and determined most appropriately at trial on their own facts and upon all appropriate and necessary information – which cannot be available at an early preliminary stage. 

Paul Lambert, BA, LLB, LLM, CTMA is adjunct lecturer and solicitor. He is the author of various publications on data protection, social networking and courtroom broadcasting.