On September 5, 2024, the European Union made history by signing the first-ever legally binding international treaty concerning AI alignment with human rights. The Convention on AI and human rights, democracy, and the rule of law was signed by Věra Jourová, the European Commission's Vice-President for Values and Transparency, during an informal conference of Council of Europe Ministers of Justice held in Vilnius, Lithuania.
The Framework Convention on AI represents the culmination of nearly two years of extensive collaborative effort led by the Committee on AI (CAI) that reunited a wide-ranging consortium, including scholars, industry professionals, government representatives, and civil society advocates.
The Convention aligns seamlessly with the AI Act, the world's first comprehensive AI regulation. It adopts a nuanced, risk-based approach for assessing and regulating AI systems potentially harmful for citizen’s fundamental rights, democracy, and the rule of law.
Key features of the Convention include:
- Mandating transparency for AI-generated content
- Implementing robust risk management protocols
- Establishing stringent documentation requirements for high-risk AI systems
- Prioritising the protection of human rights in AI development and deployment
- The concept of "regulatory sandboxes" to facilitate safe AI experimentation and innovation
- Enhanced accountability measures and remediation processes
The Framework Convention obtained support beyond the European Union, with several nations joining as signatories. Among these were Andorra, Georgia, Iceland, Norway, the Republic of Moldova, San Marino, the United Kingdom, Israel and the United States of America.
However, the signature expresses the European Union’s intention to become a member of the Convention. The next step will be the European Commission’s preparation of a proposal for a Council decision to conclude the Convention.