The European Parliament approved the Council of Europe’s Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law on March 12, 2026. The vote, according to reports, was 455 in favour, 101 against, and 74 abstentions. Official confirmation of the precise count has not been independently verified at the T1 level.
The convention is widely described as the first multilateral legally binding international treaty specifically dedicated to AI governance. That framing requires one clarification: the EU AI Act is legally binding, but it is regional EU law. This treaty is an international instrument of the Council of Europe, a separate body from the EU, with 46 member states including non-EU countries. The geographic scope is meaningfully different.
The Framework Convention’s stated goals include ensuring AI systems respect human rights, democracy, and the rule of law, with obligations covering transparency, accountability, and meaningful human oversight. According to the U.S. Department of State, the convention applies to both public authorities and private actors involved in AI activities, a scope that matters for multinational organizations that might otherwise assume the instrument is government-facing only.
The Parliament’s approval is not entry into force. Ratification by individual Council of Europe member states is the next step. The timeline to effectiveness is not confirmed in the current package and requires further reporting.