The European Commission, represented by Vice-President for Values and Transparency, Věra Jourová, has signed the Council of Europe Framework Convention on Artificial Intelligence, Human Rights, Democracy, and the Rule of Law. This historic agreement, signed during an informal conference of Ministers of Justice in Vilnius, Lithuania, marks the first legally binding international treaty on artificial intelligence (AI).
This Framework Convention is a significant step forward in ensuring AI technologies are aligned with fundamental rights, democracy, and the rule of law. It is designed to be compatible with European Union legislation, particularly the EU AI Act – the world’s first comprehensive AI regulation. The Convention addresses risks posed by AI, laying out a clear roadmap for managing its impact on society.
Key Elements of the Convention
The Convention reflects the EU’s strong commitment to human-centric AI, focusing on safeguarding human rights while fostering innovation. Several critical principles of the EU AI Act are embedded within the treaty, including:
- Human-centric AI: Ensuring that AI systems are developed and deployed with respect for human rights, democratic values, and the rule of law.
- Risk-based approach: Prioritising oversight and safeguards based on the potential risks AI systems pose.
- Trustworthy AI: Promoting transparency, robustness, safety, and sound data governance.
- Clear AI labelling: Requiring transparency in AI-generated content and interactions with AI systems.
- Accountability and oversight: Strengthening documentation and accountability mechanisms, with provisions for regulatory sandboxes to support safe AI innovation.
These provisions aim to create a balanced framework that manages AI risks without stifling innovation. The Convention also allows for exemptions in certain areas, such as research, development, and national security, ensuring flexibility in its implementation.
A Global Approach to AI Regulation
The Convention is open to accession not only by EU and Council of Europe member states but also by other countries around the world, making it a truly global initiative. While public authorities and private entities alike are required to adhere to the Convention’s obligations, private actors have some flexibility in applying alternative, appropriate measures that align with its goals.
The Path Ahead
The signing of the Convention signals the European Union’s intent to become a full Party to the agreement. The next step involves the European Commission preparing a proposal for a Council decision to conclude the Convention, followed by approval from the European Parliament.
Negotiations for this ground-breaking treaty began in September 2022 under the Council of Europe’s Committee on AI (CAI) and involved a wide range of stakeholders, including civil society, academia, industry, and international organisations. This inclusive process has helped ensure the Convention addresses the diverse and complex challenges posed by AI while safeguarding fundamental rights and promoting global cooperation.
The EU’s leadership in AI regulation is once again at the forefront, setting the stage for a future where AI development is not only innovative but also ethical and human-centric.
