Speed read

Controversial investigative journalist, Ian Wishart, wrote a book, Breaking Silence, about the twins’ mother, Macsyna King.  She collaborated with him for the book, against a background of suggestions in the public arena implicating her in relation to the deaths, despite a coroner’s report that the twins died while in the father’s sole care.

Christopher Murray established an FB page called “Boycott the Macysna King Book”.  The page got 250,000 comments by third parties, some said to be defamatory of Ian Wishart. Could Mr Murray be liable for any defamatory comments  by third parties on the page he set up? Yes, said the Court of Appeal,1    but with important reservations, one of which  is that each online liability situation cannot be pigeon- holed as factual scenarios vary.

A takeaway, at least as to FB pages, is that there is liability only where there is actual knowledge of the defamatory statement: remove potentially defamatory comments ASAP.  This might apply to other online scenarios too but it is at least possible that mere knowledge that there is likely to be defamatory statements is enough (requiring greater vigilance in circumstances that pose the prospect of defamatory statements).

This is the first major NZ case on liability online for third party statements. It moves toward greater clarity, but there is a way to go.

It’s a valuable decision for Australia, the UK and Canada too.

It also touched on how a writer might be liable when she puts in a link to another website (a  hyperlink) which is defamatory.  There are indications that merely linking the defamatory piece won’t lead to liability but that changes if say, the writer adopts or endorses the piece.  However, this is not yet resolved here. The takeaway is to still check linked pieces if they are in potentially risky territory.

The Detail

We’ve been writing on the evolution of online liability for third party statements for some years, such as here. This case is the biggest step so far. The legal background is that, in the hard copy world, those involved in publication, from the bookseller to the publisher, can all be liable for the defamation, just as a radio station can be liable for defamatory statements by live Talk Back callers. However, if “innocent dissemination” can be shown, the party avoids liability.

The trick is to apply that law, built up over many years, to the new online environment. Will the host of the FB page be liable for third party comments on his page? Will Facebook (or the ISP) be liable? (We’ll write about that, again, later, as it is not involved in this case).

\In the most comprehensive analysis we’ve seen internationally, going through various fact scenarios and precedents, the Court of Appeal showed how difficult it is to draw parallels from the hard copy world. In the end, policy considerations became important (including issues as to freedom of speech).

One of the challenges is that two step process outlined above:

  1. Who in the chain is liable (here, the FB host for third party comments) and how?
  2. Can the liable person show innocent ?

​​The Court dealt with that, against the  important background that the statutory dissemination defence does not extend beyond more traditional parties (eg booksellers and newspaper deliverers) to FB page hosts. (That is something that calls for statutory reform.)

The host of an FB page can delete comments, and in fact Mr Murray did delete around 50 of them, but then it became hard with such big volumes.

In what circumstances would the FB host be liable? There were two options, said the Court:2

  1. the “actual knowledge” test: “they know of the defamatory statement and fail to remove it within a reasonable time in circumstances that give rise to an inference that they are taking responsibility for it”.
  2. the “ought to know” test: “they do not know of the defamatory posting but ought, in the circumstances, to know that postings are being made that are likely to be defamatory.”

The “ought to know” captures a lot more and, as the Court said, runs contrary to the view that defamation is a head of claim that is based on intention.

It can make a big difference. Here for example, Mr Murray could be liable for all of the defamatory statements in the quarter of a million (“ought to know” test) or only those he knows about (“actual knowledge” test).

So, having analysed cases, fact scenarios and policy, the Court landed on limiting liability to situations where the host has only “actual knowledge”. The host must actually know of the statement (and not removed it within a reasonable time).

Although the specific facts in each case dominate and can change things, certainty for FB users was one reason for this, said the Court:3

“the ought to know test is uncertain in its application. Given the widespread use of  Facebook, it is desirable that the law defines the boundaries with clarity and in a manner that Facebook page hosts can regulate their activities to avoid unanticipated risk.”

​A takeaway, at least as to FB pages, is to remove potentially defamatory comments ASAP. There is less need to proactively monitor, at least on FB pages. This might apply to other online scenarios too, where, additionally, it is possible the “ought to know” test might apply (requiring greater vigilance in circumstances that pose the prospect of defamatory statements).

Hyperlinks to defamatory content

The judgment touches on liability where a writer includes a hyperlink to a defamatory online statement. This wasn’t resolved and so the law is yet to be decided. Until then, writers should take a conservative line as  to hyperlinked material even though there are signs the liability will be limited. For example, reference was made to what the Canadians have said in their highest court.4 Essentially:

  • the starting point is that the hyperlink is just a reference and does not communicate content (so there is no liability);
  • but the position may be different where the hyperlinker presents the hyperlinked content in a way that repeats it (eg by adoption or endorsement of the content of the hyperlinked text).