Ursula Cheer, Professor of Law, University of Canterbury, Christchurch comments on defamation, liability, publication, the internet and the recent HC case: A v Google.

Many questions have been raised about the issue of publication on the internet and whether it should be treated any differently than publication in any other media, such as hard copy newspapers, broadcasting, on bill boards and the like. There is at least no doubt that publication on the internet per se is publication for defamation purposes. In O'Brien v Brown [2001] DCR 1065 the court would not accept an argument that the "culture" of the internet allows more robust comment than elsewhere, and that trading insults is acceptable there. The judge in that case could not see why the developing culture in the use of such communication on the internet would entitle users to  greater freedom of expression than in any other method of publication.

What is defamatory publication on a website?

It has also been indicated in New Zealand some time ago in International Telephone Link Pty Ltd v IDG Communications Ltd  H C Auckland, 19 January 1998, CP 344/97 that referring in one publication to another's website can amount to publication of defamatory material to be found on that website. However, doubt may have been cast on that approach because in Crookes v Newton [2011] 3 SCR 269, the Supreme Court of Canada refused to find that the inclusion of deep and shallow hyperlinks in an online article to websites containing allegedly defamatory material was publication of the material on those websites. The majority in Crookes reached this conclusion by likening hyperlinks to footnotes.

Determining publication

In Dow Jones v Gutnick 194 ALR 433 the High Court of Australia determined that the place material is published online is where that information is downloaded. Other problems being addressed in various jurisdictions include how you prove something  has been published on the internet – some judges are using hits on a website, which is pretty rough and ready; and how long something can stay on the internet to give rise to a successful claim.  In general,  there is ongoing uncertainty about these issues.  

Who is liable?

One difficult question, still unresolved, is how far internet service providers (ISPs) are liable for defamation on the internet, by providing access for subscribers. Google has been to the fore recently.  In the UK, Tamiz v Google Inc [2012] EWHC 449, for example, involved a claim arising from a statement posted by a blogger on the popular Blogger platform on a blog called the London Muslim blog. The bloggers involved were not proceeded against.  The comments contained abuse but also serious allegations of criminality. The claimant complained to Google UK using an online ‘report abuse’ function, which passed the matter on to the USA, and Google Inc were in the process of dealing with the complaint when the blogger removed the material himself.

Owner v publisher

The judge noted that there is as yet no definitive decision as to how web publishers fit into the traditional common law view of publication. Google will on principle not remove offending material once notified. Instead, the company contacts the blogger concerned.  The judge in Tamiz noted that the Blogger platform contains more than half a trillion words and 250,000 new words are added every minute! The judge did not think that notification or the ability to remove material should automatically make Google a publisher. He thought much depended on the facts in each case and decided that here, Google was like an owner of a wall covered in defamatory graffiti where failure to remove does not necessarily make the owner a publisher.  Owners like this are seen as purely passive, and so are not publishers. This goes against some previous law, which has suggested that liability should follow where a provider has and oversees an content policy and will have actual knowledge of possible defamatory material being carried at various times. 

Is a search engine a publisher?

More recently, the novel question has arisen whether search engines themselves can be publishers.  What is at issue in these cases is whether Google can be held responsible for delivering search results when the names of certain individuals are ‘googled’, the results usually taking the form of reference to defamatory material and a link to third party websites where that information is located. In one English case, Metropolitan International Schools Ltd v Designtechnica Corp (2009 EWHC 1765 (QB))  it has been held Google was not a publisher when delivering such results.

A v Google New Zealand Ltd

The issue has now been raised in a New Zealand case  in the High Court - A v Google New Zealand Ltd HC Auckland, CIV: 2011-404-002780, 12 September 2012, this year. Here the claimant argued he was defamed by the publication of results which pointed to a website in the United States containing allegedly defamatory statements. The High Court noted the matter was novel and therefore declined to deal with it summarily – ie:  without a full trial. But the Court did express some interesting views about the issue.

The problem is that when a search is conducted, the search engine runs an automated search process on the keywords according to the pre-programmed algorithms, and produces a list of results. The results are in the form of a list of hyperlinks to web pages in order assessed by relevance. Under each link is a small excerpt or “snippet” from the web page that demonstrates that page's relevance to the search terms. Search engines can be programmed to block a specific web page. However, even if this is done, the web page is still accessible on the internet. Also, the same content can still appear on another URL - search engines have no control over web page creators, who determine their own content and can move it from one web page to another. So blocking a link will not necessarily effectively get rid of any offending material. 

The plaintiff in this case lost their application because they could not establish that Google New Zealand Ltd, which had been pursued as defendant, had any ability to control publication, even though it had passed on the plaintiff’s complaint to Google Inc, which is incorporated in the USA.

The judge did have sympathy for the plaintiff's position, however. The creator of the main website containing the defamatory material was apparently based in the United States, and has a policy of not removing any material even if requested. The defamatory material was posted by an anonymous person, who apparently lived New Zealand. The judge did not think that person would be tracked down.

Questioning automation, nature of publication and freedom of expression 

Even though he rejected the plaintiff’s claim, the judge went on to look at the situation if it could be argued that Google NZ was sufficiently connected to the search results to be considered a publisher of them.  Would it make any difference that the search results are generated by an automated system, without judgement as to the content of the information generated? The defendant, of course, argued that a finding of liability would result in  a disproportionate chilling effect on internet speech, in breach of s 14 of the NZ Bill of Rights. They also relied on the Canadian case mentioned, Crookes v Newton.

The plaintiff argued, amongst other things, that search engines are not mere conduits like ISPs, telephone carriers and  deliverers of snail mail. This is because they intend to publish search results. Search engines have deliberately chosen to return “snippets” to make their product more amenable to users and thus increase revenues. So it was argued that it would be better to treat search engine operators as publishers, but with access to a defence like the existing innocent dissemination defence which allows those who have no knowledge of carrying defamatory content and who have not been negligent to escape liability.

How far is a search engine responsible?

In A v Google, Associate Judge Abbott considered the complex issues of law would have to be decided in a proper factual context at some later stage. But he also gave some clues as to what might be relevant to that opportunity when it does come before a New Zealand court. The judge stated there may be a need to consider whether there has been ‘a stamp of human intervention’ in each case, that the way the search engine is written will have to be addressed also and the public policy concerns must be addressed.

Responding to the argument raised about chilling freedom of expression, the judge said that whilst the right to freedom of expression in the NZBORA has to be considered when developing the law in this area, it may not in fact be an unreasonable limit upon that right to hold that a search engine is a publisher of both specific URLs and words that appear in snippets. He attached weight here to the fact that these will be words search engine providers have chosen to include, thus taking hyperlinks beyond the status of mere footnotes. 

Judge Abbott thought it possible that search engine providers might even be responsible where an offending hyperlink is deactivated but the snippet continues to appear because that is still communication of a statement to just one other person as required by defamation law.  Ultimately, the judge appeared of the view that excluding from the definition of publication the repetition of information where repetition occurs without human input could unnecessarily confuse this area of the law.

The judgment indicates a reluctance to create exceptions to the previous law for the internet, and a preference for holding a search engine liable where appropriate, with access to the defence of innocent dissemination or other appropriate defences. Although a substantive decision has been left for another day, it may be that we simply cannot generalise about the internet and each case must be tried using the general principles that have already been established prior to the advent of the internet.  There has never been liability for accidental publication.  But publication in the internet is not in that category.  Levels of knowledge and control do vary depending on the service offered.  Those features cannot be ignored in the assessment of whether ISPs should be liable in defamation.

 Article by  Ursula Cheer, Professor of Law, University of  Canterbury,  Christchurch