Europe’s top court sets new line on policing illegal speech online

Europe’s top court has set a new line for the policing of illegal speech online. The ruling has implications for how speech is regulated on online platforms — and is likely to feed into wider planned reform of regional rules governing platforms’ liabilities.

Per the CJEU decision, platforms such as Facebook can be instructed to hunt for and remove illegal speech worldwide — including speech that’s “equivalent” to content already judged illegal.

Although any such takedowns remain within the framework of “relevant international law”.

So in practice it does not that mean a court order issued in one EU country will get universally applied in all jurisdictions as there’s no international agreement on what constitutes unlawful speech or even more narrowly defamatory speech.

Existing EU rules on the free flow of information on ecommerce platforms — aka the eCommerce Directive — which state that Member States cannot force a “general content monitoring obligation” on intermediaries, do not preclude courts from ordering platforms to remove or block illegal speech, the court has decided.

That decision worries free speech advocates who are concerned it could open the door to general monitoring obligations being placed on tech platforms in the region, with the risk of a chilling effect on freedom of expression.

Facebook has also expressed concern. Responding to the ruling in a statement, a spokesperson told us:

“This judgement raises critical questions around freedom of expression and the role that internet companies should play in monitoring, interpreting and removing speech that might be illegal in any particular country. At Facebook, we already have Community Standards which outline what people can and cannot share on our platform, and we have a process in place to restrict content if and when it violates local laws. This ruling goes much further. It undermines the long-standing principle that one country does not have the right to impose its laws on speech on another country. It also opens the door to obligations being imposed on internet companies to proactively monitor content and then interpret if it is “equivalent” to content that has been found to be illegal. In order to get this right national courts will have to set out very clear definitions on what ”identical” and ”equivalent” means in practice. We hope the courts take a proportionate and measured approach, to avoid having a chilling effect on freedom of expression.”

The legal questions were referred to the CJEU by a court in Austria, and stem from a defamation action brought by Austrian Green Party politician, Eva Glawischnig, who in 2016 filed suit against Facebook after the company refused to take down posts she claimed were defamatory against her.

In 2017 an Austrian court ruled Facebook should take the defamatory posts down and do so worldwide. However Glawischnig also wanted it to remove similar posts, not just identical reposts of the illegal speech, which she argued were equally defamatory.

The current situation where platforms require notice of illegal content before carrying out a takedown are problematic, from one perspective, given the scale and speed of content distribution on digital platforms — which can make it impossible to keep up with reporting re-postings.

Facebook’s platform also has closed groups where content can be shared out of sight of non-members, and where an individual could therefore have no ability to see unlawful content that’s targeted at them — making it essentially impossible for them to report it.

While the case concerns the scope of the application of defamation law on Facebook’s platform the ruling clearly has broader implications for regulating a range of “unlawful” content online.

Specifically the CJEU has ruled that an information society service “host provider” can be ordered to:

• … remove information which it stores, the content of which is identical to the content of information which was previously declared to be unlawful, or to block access to that information, irrespective of who requested the storage of that information;
• … remove information which it stores, the content of which is equivalent to the content of information which was previously declared to be unlawful, or to block access to that information, provided that the monitoring of and search for the information concerned by such an injunction are limited to information conveying a message the content of which remains essentially unchanged compared with the content which gave rise to the finding of illegality and containing the elements specified in the injunction, and provided that the differences in the wording of that equivalent content, compared with the wording characterising the information which was previously declared to be illegal, are not such as to require the host provider to carry out an independent assessment of that content;
• … remove information covered by the injunction or to block access to that information worldwide within the framework of the relevant international law

The court has sought to balance the requirement under EU law of no general monitoring obligation on platforms with the ability of national courts to regulate information flow online in specific instances of illegal speech.

In the judgement the CJEU also invokes the idea of Member States being able to “apply duties of care, which can reasonably be expected from them and which are specified by national law, in order to detect and prevent certain types of illegal activities” — saying the eCommerce Direction does not stand in the way of states imposing such a requirement.

Some European countries are showing appetite for tighter regulation of online platforms. In the UK, for instance, the government laid out proposals for regulating a board range of online harms earlier this year. While, two years ago, Germany introduced a law to regulate hate speech takedowns on online platforms.

Over the past several years the European Commission has also kept up pressure on platforms to speed up takedowns of illegal content — signing tech companies up to a voluntary code of practice, back in 2016, and continuing to warn it could introduce legislation if targets are not met.

Today’s ruling is thus being interpreted in some quarters as opening the door to a wider reform of EU platform liability law by the incoming Commission — which could allow for imposing more general monitoring or content-filtering obligations, aligned with Member States’ security or safety priorities.

“We can trace worrying content blocking tendencies in Europe,” says Sebastian Felix Schwemer, a researcher in algorithmic content regulation and intermediary liability at the University of Copenhagen. “The legislator has earlier this year introduced proactive content filtering by platforms in the Copyright DSM Directive (“uploadfilters”) and similarly suggested in a Proposal for a Regulation on Terrorist Content as well as in a non-binding Recommendation from March last year.”

Critics of a controversial copyright reform — which was agreed by European legislators earlier this year — have warned consistently that it will result in tech platforms pre-filtering user generated content uploads. Although the full impact remains to be seen, as Member States have two years from April 2019 to pass legislation meeting the Directive’s requirements.

In 2018 the Commission also introduced a proposal for a regulation on preventing the dissemination of terrorist content online — which explicitly included a requirement for platforms to use filters to identify and block re-uploads of illegal terrorist content. Though the filter element was challenged in the EU parliament.

“There is little case law on the question of general monitoring (prohibited according to Article 15 of the E-Commerce Directive), but the question is highly topical,” says Schwemer. “Both towards the trend towards proactive content filtering by platforms and the legislator’s push for these measures (Article 17 in the Copyright DSM Directive, Terrorist Content Proposal, the Commission’s non-binding Recommendation from last year).”

Schwemer agrees the CJEU ruling will have “a broad impact” on the behavior of online platforms — going beyond Facebook and the application of defamation law.

“The incoming Commission is likely to open up the E-Commerce Directive (there is a leaked concept note by DG Connect from before the summer),” he suggests. “Something that has previously been perceived as opening Pandora’s Box. The decision will also play into the coming lawmaking process.”

The ruling also naturally raises the question of what constitutes “equivalent” unlawful content? And who and how will they be the judge of that?

The CJEU goes into some detail on “specific elements” it says are needed for non-identical illegal speech to be judged equivalently unlawful, and also on the limits of the burden that should be placed on platforms so they are not under a general obligation to monitor content — ultimately implying that technology filters, not human assessments, should be used to identify equivalent speech.

From the judgement:

… it is important that the equivalent information referred to in paragraph 41 above contains specific elements which are properly identified in the injunction, such as the name of the person concerned by the infringement determined previously, the circumstances in which that infringement was determined and equivalent content to that which was declared to be illegal. Differences in the wording of that equivalent content, compared with the content which was declared to be illegal, must not, in any event, be such as to require the host provider concerned to carry out an independent assessment of that content.

In those circumstances, an obligation such as the one described in paragraphs 41 and 45 above, on the one hand — in so far as it also extends to information with equivalent content — appears to be sufficiently effective for ensuring that the person targeted by the defamatory statements is protected. On the other hand, that protection is not provided by means of an excessive obligation being imposed on the host provider, in so far as the monitoring of and search for information which it requires are limited to information containing the elements specified in the injunction, and its defamatory content of an equivalent nature does not require the host provider to carry out an independent assessment, since the latter has recourse to automated search tools and technologies.

“The Court’s thoughts on the filtering of ‘equivalent’ information are interesting,” Schwemer continues. “It boils down to that platforms can be ordered to track down illegal content, but only under specific circumstances.

“In its rather short judgement, the Court comes to the conclusion… that it is no general monitoring obligation on hosting providers to remove or block equivalent content. That is provided that the search of information is limited to essentially unchanged content and that the hosting provider does not have to carry out an independent assessment but can rely on automated technologies to detect that content.”

While he says the court’s intentions — to “limit defamation” — are “good” he points out that “relying on filtering technologies is far from unproblematic”.

Filters can indeed be an extremely blunt tool. Even basic text filters can be triggered by words that contain a prohibited spelling. While applying filters to block defamatory speech could lead to — for example — inadvertently blocking lawful reactions that quote the unlawful speech.

The ruling also means platforms and/or their technology tools are being compelled to define the limits of free expression under threat of liability. Which pushes them towards setting a more conservative line on what’s acceptable expression on their platforms — in order to shrink their legal risk.

Although definitions of what is unlawful speech and equivalently unlawful will ultimately rest with courts.

It’s worth pointing out that platforms are already defining speech limits — just driven by their own economic incentives.

For ad supported platforms, these incentives typically demand maximizing engagement and time spent on the platform — which tends to encourage users to spread provocative/outrageous content.

That can sum to clickbait and junk news. Equally it can mean the most hateful stuff under the sun.

Without a new online business model paradigm that radically shifts the economic incentives around content creation on platforms the tension between freedom of expression and illegal hate speech will remain. As will the general content monitoring obligation such platforms place on society.