Font size:
Print
Council of Europe’s Convention on Artificial Intelligence (AI)
Context:
The United States, the European Union, and the United Kingdom (UK) are anticipated to sign the Council of Europe’s Convention on Artificial Intelligence (AI).
The Council of Europe, established in 1949, is an international organisation separate from the EU with a mandate to safeguard human rights. It has 46 member countries, including all 27 EU member states.
More on news
- This treaty emerges at a time when various governments and smaller multilateral groups are introducing new regulations and agreements to regulate AI tools.
- These include the G7’s AI pact (October 2023), Europe’s AI Act, and the Bletchley Declaration signed by 28 countries (November 2023).
About the treaty
- This will be the first “legally binding” international treaty governing the use of this transformative technology.
- The treaty, developed over the past two years by more than 50 nations, adopts a risk-based approach to the design, development, use, and decommissioning of AI systems.
- It applies to AI use in both the public and private sectors, including companies acting on behalf of the public sector, and will be enforced globally.
- Signatory countries will be held responsible for “any harmful and discriminatory outcomes of AI systems” and must ensure that “AI outputs respect equality and privacy rights, while providing legal recourse for victims of AI-related rights violations.”
- Officially named the “Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy, and the Rule of Law”, the treaty was opened for signatures during a conference of the Council of Europe.
Obligations and Applicability
Upon ratification, member states will be required to:
- Ensure their AI systems are “consistent with obligations to protect human rights.”
- Prevent the use of AI systems to “undermine the integrity, independence, and effectiveness of democratic institutions and processes,” including upholding the separation of powers, judicial independence, and access to justice.
- Implement measures to “safeguard democratic processes throughout the AI lifecycle,” ensuring fair access to public debate, the ability to participate, and the freedom to form opinions.
- Certain exemptions apply under the Framework Convention, such as those related to national security and research and development.
Issues and Concerns
- While the treaty is hailed as “legally binding,” concerns have been raised about its lack of provisions for punitive sanctions, such as penalties or fines.
- Enforcement primarily relies on “monitoring,” which some see as insufficient in deterring non-compliance.