Something Different for the Midweek: Google and Criminal Liability

Yesterday Pangloss was very happy to have a guest lecture for her Internet Law class given by Trevor Callaghan, Managing Product Counsel of Google UK. Trev gave a hilarious lecture on the law relating to search and copyright, which conbined legal insight, practical tips, and social responsibility with some Glasgow humour that would have put Armando Iannuci of The Thick Of It fame to shame (albeit with (slightly) less swearing). I enjoyed it, lots, and i think the students did too.

Anyway, this all reminded me that actually quite a few things are going on I should be talking about as well as (or perhaps even in combination with) the Digital Economy Bill. One of these, which has received suprisingly little press (even wonderful OUT-LAW hasn't mentioned it since February) , is that right now, four Google executives - including Privacy CEO Peter Fleischer- are on trial - yes, criminal trial - in Italy, in relation to a short phonecam video made by some school children of a bullying incident involving a child with learning disabilities, and then posted on Google Video.

In Italy, it appears that libel and , possibly, infringement of privacy laws, can be a matter of criminal as well as civil law. Google took down the video on notice within a day of receiving an official complaint from a consumer group, although the video had been online for about 2 months before that. Italian prosecutors investigated for two years but then decided to proceed.

For Pangloss this seems a not very difficult case that ought to be easily decided under the EC E-Commerce Directive safe harbours in Art 14 and 15, as often discused in this blog. If these aren't implemented into Italian law, then it would seem Italy must be in breach of EC law itself. Google was clearly a host here, and Art 14 provides that such sites are protected from criminal liability for the activity of users of the service, unless they receive actual notice, and fail to take down expediently. This is a case about criminal liability so there is no need even to move to the second branch of Art 14 (which is far more controversial) and discuss whether Google should have known - ie had constructive knowledge - of the activity or content. Injunctions would have been relevant, despite the safe harbours, but these are not the issue as Google already took down straightaway on notice.

So why on earth is this case coming to trial? Pangloss is perplexed. One possibility as noted above is that simply that Italy's domestic law is in breach of EC law (in which case Google should have a Francovich claim for damages against the Italian government, though that may not be much comfort to the men awaiting trial.) Another possibility, though rather an unlikely one, is that the Italian prosecutors have confused the activities of Google as a search engine, with Google as a host. The ECD does not give search engines , or hyperlinkers , a special immunity from liability as it does hosts and "mere conduits" : though a number of EC countries have in fact decidd to extend such an immunity, either under Art 12 or 14, or both. However in this case case it seems pretty clear Google was a host not a hyperlinker in terms of liability. So, what on earth quid iuris?

Another remote possibility is that the suggestion is that Google as a provider of free services does not gain the benefit of the Art 14 safe harbour. This uncertainty has been around for a long time, since only providers of "information society services"(ISSPs) get the benefit of Arts 12-15 and that definition is of an online service "normally provided for remuneration" (see recitals 17 and 18). Yet majority opinion has long felt that this particular point is no obstacle to the likes of Google (or Facebook, or Hotmail?) claiming safe harbours.

First, while renumeration might not come directly from users, it certainly does come in the form of the adverts Google place alongside its services. Second, search services are certainly something that would "normally" be paid for if they weren't, happily, often provided for free: they are of huge commercial value . Thirdly, it seems a strange policy in terms of public interest which would discriminate against services of great public value provided for free, in favour of those given purely for direct consideration.

There is no clear ECJ ruling on this yet but there is likely to be soon: in the upcoming Adwords conjoined referrals to the ECJ (Google France v Louis Vuitton, etc), the Advocate-General has already given a preliminary opinion in which he found:

"There is nothing in the wording of the definition of information society services to exclude its application to the provision of hyperlinks and search engines, that is to say, to Google’s search engine and AdWords. The element ‘normally provided for remuneration’ may raise some doubts as regards Google’s search engine, but, as has been pointed out, the search engine is provided free of charge in the expectation of remuneration under AdWords. (68) Since both services are also provided ‘at a distance, by electronic means and at the individual request of the recipient of services’, they fulfil all the requirements necessary to be regarded as information society services."(para 131)

And for what it is worth, a roughly similar finding was reached, albeit obiter and with an admission of some possibility of doubt , in the recent English libel case of Metropolitan v Designtechnica, where Eady J opined: "it would appear on balance that the provisions of the 2002 Regulations [defining an ISSP] are apt to cover those providing search engine services." (para 84)

So what does that leave? Well there is perhaps a clue in the New York Times account.

"Google and the prosecutors agree the video was uploaded Sept. 8 and removed Nov. 7, 2006. The prosecutors presented evidence showing that in early October, a month before the video’s removal, there were comments posted saying that it should be taken down. One of those messages read, “This is shameful! This should be taken down immediately.”

“It is reasonable to imagine that comments like this were followed by requests by these same people that the video be removed,” the prosecutors wrote in the document they presented to the judge."

So when are such shocked responses or "requests", "actual notice" as required by Art 14? Do comments on a video hosting site cut it, as opposed to an official request for takedown? To put it another way: does a hosting service have a duty to read comments about videos posted by, and probably of interest only to, their creators and viewers? Surely not.

Compare the situation to the original world Art 14 was designed to deal with, that of web 1.0. If Demon Internet hosted a basic site for (let's say) Anglers Magazine, and it contained a chatroom where libellous remarks were made about particular fly-fishers, would Demon be expected to monitor that chatroom for explicit or implied requests to take down those comments? Again, surely not. It would be up to the aggrieved angler to send his request for take down direct to Demon. The whole point of Art 14 was to reassure host providers they had no need to monitor the activities of those to whom they provided hosting services. Not only would this involve huge expenditure of effort and cost, but it might also be privacy invasive and chilling of free speech. Art 15 states this absolutely explicitly:

"Member States shall not impose a general obligation on providers, when providing the services covered by Articles 12, 13 and 14, to monitor the information which they transmit or store, nor a general obligation actively to seek facts or circumstances indicating illegal activity."

Still another way to put this is to ask , what are the minimum requirements for notice? This is a perennial problem. The US DMCA largely gets it right, with a statutory form which requires a complainant to give clear details including their own address and status as rightsholder, and provides sanctions for false accusations. The ECD, being a EC wide framework, is hopelessly vague. The UK's own regs help a little but not much - there is no DMCA type statutory notice but Reg 22 of our E Commerce Regulations does state that

"In determining whether a service provider has actual knowledge ... a court shall take into account all matters which appear to it in the particular circumstances to be relevant [including] whether a service provider has received a notice through a means of contact made available in accordance with regulation 6(1)(c)" - ie, their official contact email address .

This stuff should be simple law (compared at least to issues like eBay and Louis Vuitton, Google and AdWords) but even it is not. The ECD deperately needs revised to get a few simple things right and harmonised across Europe: what form should "actual notice" take; what does "expediently" mean; what is constructive notice; when, if ever, can an obligation to filter proactively be placed on ISSPs; what immunities should search engines (and hyperlinkers and aggregators) have. Pangloss loves this stuff but even she is tired of writing the same stuff over and over again. It is time to review the ECD.

PS and in the interest of public policy but with just a hint of minx-itude, I have helped draft a proposed amendment to the Digital Economy Bill for ORG which would aim to clarify some of these very matters, at least for the UK. See you in the House of Lords! :-)

New UK Internet Libel case coming?

Thanks for the heads up from @loveandgarbage (c/o Twitter) for the following news from David Osler, author of Dave's Part, who is facing libel action from Tower Hamlets Tory activist Johanna Kaschke, following a post on this blog in 2007. She is also (says Dave) suing two other Labour Party members, Alex Hilton and John Gray, over related issues.

"The uncontested facts here are that Ms Kaschke, as a student and member of the centre-left SPD in her native Germany in the 1970s, helped to organised a benefit concert for Rote Hilfe, an organisation officially designated left-extremist, designed to raise funds for the legal fees of Baader-Meinhof Gang suspects; that she was herself subsequently arrested on suspicion of terrorism; and that she spent several months on remand, after which she was released and compensated for unfair imprisonment.

It is further uncontested that Ms Kaschke nominated herself as Labour candidate for Bethnal Green & Bow in 2007; that she received just one vote; that shortly thereafter she defected to George Galloway's Respect party; shortly after that, she joined an as-yet-unspecified Communist Party; and that shortly after that, she became a Conservative. She was, in other words, a member of four political parties in 12 months.

Interestingly, the jury will be asked to rule on whether or not it is libellous to call somebody 'one cherry short of a Schwarzwalderkirschtorte'. Not my words, but those of a reader, left in the comments box. If I lose on that point, the consequences for internet freedom of speech are clearly considerable."

Remarkable, and not just for the linguistics (German cookery experts are invited to elucidate me on the exact translation).

Sites such as the BBC , Yahoo! etc have of course long worried about the liability implications of comment sections on "live" blogs whether moderated or unmoderated. It seems well settled that a host site owner can be liable for publishing the defamatory remarks of commenters. It is equally well settled however that under both Defamation Act 1996 s 1 and the EC E-Commerce Directive Arts 14 and 5, defences are open to hosts in respect of content posted by another.

The interesting point here is I think about the quality of the quoted comment. The English courts have so far, rather admirably, taken a fairly robust attitude towards too quickly attributing the quality of libel to remarks made in the typical hasty cut and thrusts of Internet babble. Most notably in Sheffield Wednesday v Hargreaves [2007] EWHC 2375: particularly

It seems to me that some of the postings which concern the Claimants border on the trivial, and I do not think that it would be right to make an order for the disclosure of the identities of users who have posted messages which are barely defamatory or little more than abusive or likely to be understood as jokes. That, it seems to me, would be disproportionate and unjustifiably intrusive. The postings which in my judgment fall into this category are those numbered 4 ("xdanielx"), which is only capable of being argued to be defamatory by devising a frankly implausible meaning, 7 ("Foot04"), which is barely if at all defamatory of the Second Claimant, 8 ("southy") and 14 ("cbrbob"), both which in my view are plainly intended as jokes and would have been unlikely to be taken seriously, let alone understood in the senses for which Mr Eardley argued, and 10 and 11 ("paulrs") which I regard as no more than saloon-bar moanings about the way in which the club is managed, rather than a serious indictment of grave mismanagement. In my view the same is true of 6 ("Auckland Owl") and 12 ("danksy"), which add to the mix a smidgeon of personal abuse of a kind which I would have thought most unlikely to be taken seriously. I take a similar view of the posting numbered 2 ("DJ Mortimer"), which is no more than mildly abusive and is fairly plainly comment.

1. The postings which I regard as more serious are those which may reasonably be understood to allege greed, selfishness, untrustworthiness and dishonest behaviour on the part of the Claimants. In the case of those postings, the Claimants' entitlement to take action to protect their right to reputation outweighs, in my judgment, the right of the authors to maintain their anonymity and their right to express themselves freely,

Thuis was of course however an action for discloure of the identity of the commenters by Norwhich Pharmacal orders , not an actual action on liability.

But see also Smith v ADFN , misreported as Adven in Edwards and Waelde 3rd ed, oh dear.

Here Eady J took possibly the firmest imaginable stance in discouraging libel actions re mere "vulgar abuse " (eg a description of plaintiff in comments as "a destructuive twerp")

The question on which I need to focus next is whether there has been persistence with regard to claims that can properly be characterised as "totally without merit". If so, I could come to the corresponding conclusion that his present application to lift the stays would to that extent also be totally without merit. I have rehearsed above a number of examples where claims have been made in respect of postings which are so obviously, in their context, either mere vulgar abuse or fair comment (sometimes both). There are also examples of a converse and corresponding interest in the subject-matter, for various reasons, such as to give rise to occasions of qualified privilege.

1. I referred to common themes in the postings, such as that of "bullying" other users and making "threatening demands" for money. That is classic fair comment territory and, in the light of the modern authorities, it is inconceivable that a jury would find any of those who expressed such a view "malicious" – let alone all of them. Opinions may be expressed in exaggerated and strident terms; the only requirement is that they be honestly held. It is fanciful to suppose that any of these people did not believe what they were saying. Even if they reached their conclusions in haste, or on incomplete information, or irrationally, the defence would still avail them. It would be wasteful to let proceedings go forward merely on the footing of a series of formulaic assertions to the effect that the individual concerned did not honestly believe what he was saying. There is accordingly no realistic prospect of any such claims achieving the only legitimate goal of vindicating reputation.
2. I would not suggest for a moment that blogging cannot ever form the basis of a legitimate libel claim. I am focusing only on these particular circumstances. It does seem to me appropriate to characterise these claims as totally without merit. I will therefore make an extended civil restraint order, which means that Mr Smith cannot launch any further libel proceedings arising out of the Langbar matter based upon bulletin board blogs without obtaining my written permission."

These remarks are technically obiter in relation to Osler's case as the issue in Smith was the maintenance of injunctions rather than a finding of libel. However in relation to the comment post in particular, in short, I do not think Mr Osler has much to worry about :-)

Internet Libel (not "liable") or Who's the Daddy(place)?

A story I meant to mention from last week - the Telegraph reported what is being called the largest ever Internet libel settlement in the UK, in relation to allegations on a site called "Dadsplace" about Gentoo, a housing development company.

"Gentoo Ltd, formerly the Sunderland Housing Group, became the subject of an attack by "a seriously defamatory, abusive and scurrilous anonymous website at dadsplace.co.uk", according to a statement read in court by the organisation's counsel, Hugh Tomlinson QC, before Mr Justice Eady today."

Eventually after some two years of malicious attacks downloaded "millions" of times, "John Finn, the owner of rival housing firm Pallion and a former local council candidate in Sunderland ...admitted his involvement, agreeing at the time to pay £125,000 towards Gentoo’s legal costs and a total of £21,000 in compensation.. he and Pallion [then] agreed to pay Mr Walls damages of £100,000 to settle his claim for libel and harassment."

The webmasters of Dadsplace were also made subject to injunctions not to repeat the offending statements but do not seem to have been sued for actual damages.

Now interestingly the solicitors for Gentoo - Olswangs - have commented publicly on why they think the settlement was so high. Factors seem to include:
- the length of the slandering campaign - two years
- the quantity of defamatory allegations - made almost daily
- the "extensive steps to publicise the Web site and their other publications" made by Dadspace - so the damage caused to the reputation was very extensive.

They also indicate how difficult it is to investigate a campaign of anonymous libel eg on a bulletin board or mailing list site, involving "months of painstaking investigation involving a combination of high-tech computer forensic work and old-fashioned evidence gathering".

Finally there are some interesting thoughts on Internet libel from Ashley Hurst the Olswang lawyer involved:

"This raises the question of whether reform is required to give the Internet the same badge of respectability that is enjoyed by other forms of media, including the press (regulated by the PCC) and television companies (regulated by Ofcom). However, the Internet is of course an entirely different medium and the answer is far from straightforward, particularly given the global reach of the Internet and the many different foreign laws that can apply. Would extending the remit of Ofcom or the PCC, or developing a voluntary code of conduct, make any difference?"

Pangloss gets an awful lot of requests to provide advice on Internet libel, though she is uncertain if this is because there is so much of it, or because her article on Net defamation (from 2000!!) comes up first in Google UK if you put in "Internet libel". (Bored students may be glad to know this piece will finally be updated in the 3rd edn of Law and the Internet upcoming.)

But most the people who contact her (unike Olswangs, perhaps, who charge :-) are not the alleged victims of libel, but are websites or hosts of some kind (often charitable or one-man outfits) who suddenly receive take down notices out the blue making vague threats of legal action, and then have no idea what their legal risks are. In an Internet culture where flaming is still fairly prevalent, these hosts often feel they have no alternative but to take down, even where they have no idea what if anything illegal or actionable has been said. This is not good for freedom of speech, democracy or indeed the morale of the voluntary/charitable sector. Sabre rattling and fear of legal risk , it seems, often overwhelms common sense and resilience.

Helpfully, the SCL website as well as providing the Olswang interview, also provides some hints to websites as to when they are liable for content posted on their site by third parties.

Pangloss doesn't disagree that a voluntary code relating to offensive content on websites might be of some use for the victims of malicious allegations (though how would it be policed? the PCC model, both of jurisdiction and sanctions, does not readily transfer, she feels, and that's before we come to the fact that web content is just as likely to be uploaded abroad as in the UK.)

But she also wonders if we do not also need to do more to protect individuals and small unincorporated associations who run or host the websites from random take down notices from anyone who is a wee bit disgruntled or wants to stifle perfectly reasonable criticism or debate.

At the very least it would be good to see a responsible body - the CABxs ? ISPA? BERR? - providing some plain language guidance on line, perhaps an advice hotline, and perhaps even an adaptable form response to takedown notices which do not meet the requirements of regulation 22 of the ECD regulations. Some take down notices do not even sometims specify what ( or where) the alleged libel IS. (The title of this piece comes from one just like this Pangloss saw yesterday - where the aggrieved sender of the take down notice knew so little he had spelt "libel" as "liable".)

As`my gift to the world Pangloss may post her own typical response letter tomorrow. After I've checked it's in no way libellous:)