The First Binding International Treaty on AI Now Has the EU as a Signatory

The EU AI Act is a European Union law. This is a different, international treaty — and with the EU's signature in May 2026, it's now the first binding AI instrument that transcends any single regional bloc.

What sets it apart from the EU AI Act

The Council of Europe's Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law (the Council of Europe is distinct from the European Union — it's a broader organization with 46 member states) is an international treaty, not a single bloc's law. It opened for signature on September 5, 2024, and — unlike the EU AI Act — is explicitly open to countries that aren't Council of Europe members too.

The European Union's ratification

On May 15, 2026, during the 135th session of the Council of Europe's Committee of Ministers in Chișinău, the European Union formally deposited its instrument of ratification of the Convention — a significant step because it makes the EU, already the bloc with the world's strictest AI law, a formal party to this additional international framework specifically focused on human rights.

What the treaty requires

The Convention covers the entire lifecycle of AI systems and establishes principles of transparency, accountability, oversight, non-discrimination, and safeguards against harmful uses — a human-rights principles framework applied specifically to AI, complementary to (not redundant with) the EU AI Act's technical-operational risk classification.

Why “open to non-European countries” is the part that matters most

The Convention's most relevant feature for the rest of the world is that any country, whether or not it's a Council of Europe member, can accede to it. That makes it a potential vehicle for countries in Latin America, Africa, or Asia to adopt an already-drafted, internationally negotiated human-rights framework for AI, instead of writing their own from scratch — an alternative to building a full national law on their own.

What it means for companies with multi-jurisdiction operations

For a company already compliant with the EU AI Act, the Council of Europe Convention shouldn't add a significant new technical burden — the transparency and non-discrimination principles it requires are, for the most part, already covered by compliance with the European law. Its real relevance is geopolitical: it's the clearest signal yet that a human-rights-based model of AI governance, not just technical risk management, is being exported beyond Europe.