Press release

The Council of Europe’s Convention on AI: No free ride for tech companies and security authorities!

The Convention on AI is intended to be the first legally binding international agreement on AI. The final round of negotiations will take place in Strasbourg starting 11 March 2024. The members of the Council of Europe (including the EU member states) and non-members such as the US, Japan and Canada will also be sitting around the negotiating table. AlgorithmWatch, over 90 civil society organizations, and prominent academics are calling on the negotiating states to regulate companies’ and national security authorities’ use of AI.

Angela Müller
Angela Müller
Head of Policy & Advocacy | Executive Director AlgorithmWatch CH

The Council of Europe’s members have signed the European Convention on Human Rights, which was adopted almost 75 years ago. Now they are about to conclude another binding agreement under international law: the Convention on Artificial Intelligence (AI). This convention aims to protect human rights, democracy, and the rule of law from AI’s harmful effects.

Exceptions for the private sector and "national security"

However, the published draft shows: We might end up with a Convention that could not apply to companies – or in which states are at least free to choose whether to apply it to companies, too. “This would send a dangerous signal: The first international rulebook on AI could thus give corporations a free pass to develop and use AI according to their own interests. The negotiating states must ensure that AI serves the interests of humanity and not those of a few big corporations – all the more so as the Council of Europe is supposed to be the watchdog for human rights in Europe,” says Angela Müller, Executive Director of AlgorithmWatch CH.

Furthermore, the convention will probably not be applicable if states allow AI to be developed and used under the guise of "national security". Very different AI applications could fall under this vague term: face recognition in public spaces, AI border protection, or scanning social media profiles. All of these areas of application require the protection of fundamental rights, a protection that must not be undermined by a free pass for "national security."

In an open letter, over 90 civil society organizations from around the world and prominent academics are therefore calling on the negotiating states to cover both the public and private sectors (i.e., not to make any exceptions for private companies) and to reject blanket exceptions in the name of national security.

A small window left

Policy-makers in Strasbourg only have a small window left to correct these shortcomings and to put fundamental rights back at the core of European AI governance. The negotiators must return to the Convention on AI’s idea ofprotecting human rights, democracy, and the rule of law.