On 12 September 2018, the European Commission presented a proposal for a regulation on preventing the dissemination of terrorist content online—dubbed the Terrorism Regulation, or TERREG for short—that contained some alarming ideas. In particular, the proposal included an obligation for platforms to remove potentially terrorist content within one hour, following an order from national competent authorities.
Ideas such as this one have been around for some time already. In 2016, we first wrote about the European Commission’s attempt to create a voluntary agreement for companies to remove certain content (including terrorist expression) within 24 hours, and Germany’s Network Enforcement Act (NetzDG) requires the same. NetzDG has spawned dozens of copycats throughout the world, including in countries like Turkey with far fewer protections for speech, and human rights more generally.
Beyond the one hour removal requirement, the TERREG also contained a broad definition of what constitutes terrorist content as “material that incites or advocates committing terrorist offences, promotes the activities of a terrorist group or provides instructions and techniques for committing terrorist offences”.
Furthermore, it introduced a duty of care for all platforms to avoid being misused for the dissemination of terrorist content. This includes the requirement of taking proactive measures to prevent the dissemination of such content. These rules were accompanied by a framework of cooperation and enforcement.
These aspects of the TERREG are particularly concerning, as research we’ve conducted in collaboration with other groups demonstrates that companies routinely make content moderation errors that remove speech that parodies or pushes back against terrorism, or documents human rights violations in countries like Syria that are experiencing war.
TERREG and human rights
TERREG was created without real consultation of free expression and human rights groups and has serious repercussions for online expression. Even worse, the proposal was adopted based on political spin rather than evidence.
Notably, in 2019, the EU Fundamental Rights Agency—tasked with an opinion by the EU parliament—expressed concern about the regulation. In particular, the FRA noted that the definition of terrorist content had to be modified as it was too wide and would interfere with freedom of expression rights. Also, “According to the FRA, the proposal does not guarantee the involvement by the judiciary and the Member States' obligation to protect fundamental rights online has to be strengthened.”
Together with many other civil society groups, we voiced our deep concern over the proposed legislation and stressed that the new rules would pose serious potential threats to fundamental rights of privacy, freedom of expression.
The message to EU policymakers was clear:
- Abolish the one-hour time frame for content removal, which is too tight for platforms and will lead to over removal of content;
- Respect the principles of territoriality and ensure access to justice in cases of cross-border takedowns by ensuring that only the Member State in which the hosting service provider has its legal establishment can issue removal orders;
- Ensure due process and clarify that the legality of content be determined by a court or independent administrative authority;
- Don’t impose the use of upload or re-upload filters (automated content recognition technologies) to services under the scope of the Regulation;
- Exempt certain protected forms of expression, such as educational, artistic, journalistic, and research materials.
However, while responsible committees of the EU Parliament showed willingness to take the concerns of civil society groups into account, things looked more grim in Council, where government ministers from each EU country meet to discuss and adopt laws. During the closed-door negotiations between the EU-institutions to strike a deal, different versions of TERREG were discussed, which culminated in further letters by civil society groups, urging the lawmakers to ensure key safeguards on freedom of expressions and the rule of law.
Fortunately, civil society groups and fundamental rights-friendly MEPs in the Parliament were able to achieve some of their goals. For example, the agreement reached by the EU institutions includes exceptions for journalistic, artistic, and educational purposes. Another major improvement concerns the definition of terrorist content (now matching the narrower definition of the EU Directive on combating terrorism) and the option for host providers to invoke technical and operational reasons for non-complying with the strict one-hour removal obligation. And most importantly, the deal states that authorities cannot impose upload filters on platforms.
The Deal Is Still Not Good Enough
While civil society intervention has resulted in a series of significant improvements to the law, there is more work to be done. The proposed regulation still gives broad powers to national authorities, without judicial oversight, to censor online content that they deem to be “terrorism” anywhere in the EU, within a one-hour timeframe, and to incentivize companies to delete more content of their own volition. It further encourages the use of automated tools, without any guarantee of human oversight.
Now, a broad coalition of civil society organizations is voicing their concerns with the Parliament, which must agree to the deal for it to become law. EFF and others suggest that the Members of the European Parliament should vote against the adoption of the proposal. We encourage our followers to raise awareness about the implications of TERREG and reach out to their national members of the EU Parliament.