On July 9th, the Grand Chamber of the European Court of Human Rights, its very highest body and one of the top human rights courts in the world, held a hearing on the landmark case of Delfi, an Estonian news outlet held liable for comments posted by users.
Access has followed the case for some time, and recently intervened in favor of Delfi and the right to pseudonymity and anonymity online. For details on our intervention and the case history check out our recent blog here. In short, lower courts have failed to protect free expression, imposing liability on the website for user comments below an article, and ludicrously declaring Delfi should “prevent” all defamatory comments.
This week, arguments in court revealed the flawed assumptions and dangerous implications of this case.
Who controls comments?
In open court, Delfi’s lawyers referred to interventions, or briefs on related topics submitted by Access and many others, to make the case against liability for internet intermediaries and for free expression online, a stance Access strongly supports. In particular, they argued that the comment section of the Delfi website is merely a passive service that carries comments. Delfi is not the author of the comments, and shouldn’t be held responsible for them.
Countering, the Estonian government argued that Delfi was liable for the comments because it exercised some degree of control over the comments, and because comments affect its revenues. To prove control, they first pointed to Delfi’s content filtering system, designed to weed out any comments with “vulgar” words, and a flagging system to report inappropriate posts. The comments would then be reviewed by Delfi staff, and removed if found to be inappropriate – exactly what happened in the incident sparking this case.
This system, referred to as “notice and takedown,” is actually a long-standing mechanism meant to protect free expression online. The practice is required in the European Union’s E-commerce directive, which Delfi complied with fully. The offending comments were taken down the very same day the request was filed, yet Delfi is still being pursued in the courts.
Next, the Estonian government argued that, because the comments could not be edited or removed by the original author after being posted, Delfi retained sole control over the comments. The government took care to clarify that it did not seek to hold other providers such as Facebook or YouTube liable for comments or content posted on their services, as the original authors retained control over content on those sites, because users can edit their comments after posting.
Critiques of Estonia’s reasoning
For a number of reasons, we disagree with the Estonian government’s central assertion that Delfi is not a mere intermediary for comments posted on its website, but is responsible for their content.
First, the comments were generated by anonymous users, not employees of Delfi.
Second, the government’s differentiation between Delfi and other platforms such as Facebook and YouTube is based on flawed assumptions. Estonia appears to believe that being unable to edit content somehow means that you are no longer the author of said work. But authorship – and liability – do not depend on a website’s functionality, or its arbitrary choice to allow editing of content once it’s posted or uploaded.
What’s more, the assumptions behind this distinction quickly break down. For example, under Delfi’s systems, a comment’s author could have a comment removed by simply contacting Delfi, or flagging the comment through its notice-and-takedown system. As such, the comments are not entirely beyond the control of the author, even after they’re posted on Delfi’s site.
In sum, websites provide a forum for users to speak, express themselves, and associate. Users’ ability to edit their own comments is irrelevant to authorship, and in any case would not prevent defamatory speech, one of the primary if ill-conceived goals of lower courts in this case.
Popularity increases liability?
Further, Estonia argued that the prominence of articles on Delfi’s homepage was determined, in part, by the number of comments they received. Therefore, a popular article was visited more times, generating more ad revenue.
The reasoning behind this argument appears to be that, because comments influence Delfi’s revenue, they are a core part of its platform, and therefore Delfi should be liable for speech that it facilitates.
Just as with control, revenue generated from comments does not affect the authorship of said comments. The original poster wrote the comment, no matter who profits from it.
Strangled by financial liability, and incentivized to over-comply
The government’s proposed resolution to the problem is, in effect, to require that Delfi pre-approve every comment, especially on controversial articles. This is absolutely ludicrous – as noted by Delfi’s lawyer, on the internet anything could be “controversial.” As some person could be personally offended or affronted by anything, every single article would have to be closely monitored.
Such a requirement also places the enforcement of the law in the hands of private companies. Placing companies in the position of playing judge, jury, and executioner overexpression online, while at the same time holding them liable for these decisions, creates a perverse incentive to censor speech.
Up to a year
Access will follow this case closely, but we expect it may be up to a year before the Grand Chamber rules on Delfi AS v Estonia. Here’s hoping they strike a different path than lower courts, and instead side with the open internet and the right to free expression online. Anything less than overturning previous decisions and eliminating intermediary liability for user-generated comments could upend the internet as we know it.