EU's AI CONVENTION

• Although numerous ethical guidelines, 'soft law' tools, and governance principles are documented globally, none are binding or likely to result in a global treaty.
• Additionally, there are no ongoing negotiations for an AI treaty at the international level.
• In this context, the Council of Europe (COE) made a significant move by adopting the Framework Convention on Artificial Intelligence and Human Rights, Democracy, and the Rule of Law, also known as the 'AI Convention'.
• The COE, an intergovernmental organisation established in 1949 with 46 member states, has created this comprehensive convention to address AI governance and its connections to human rights, democracy, and the responsible use of AI. The framework convention will be open for signature on September 5, 2024.

Article 1 of the convention states: "The provisions of this Convention aim to ensure that activities within the lifecycle of artificial intelligence systems are fully consistent with human rights, democracy, and the rule of law."

Article 3 further elaborates: "The scope of this Convention covers the activities within the lifecycle of artificial intelligence systems that have the potential to interfere with human rights, democracy, and the rule of law as follows:

• The exemptions in Articles 3.2, 3.3, and 3.4 of the AI convention are broad, addressing the protection of national security interests, research, development and testing, and national defence.
• Consequently, military applications of AI are not covered by the convention. While this exclusion is concerning, it is a pragmatic decision given the current lack of consensus on regulating military AI applications.
• The exemptions in Articles 3.2 and 3.3, though broad, do not entirely exclude the convention's applicability regarding national security and testing.
• The 'General Obligations' of the convention focus on the protection of human rights (Article 4), the integrity of democratic processes, and respect for the rule of law (Article 5).
• Although disinformation and deep fakes are not specifically addressed, parties to the convention are expected to take measures against them under Article 5.
• Furthermore, Article 22 indicates that parties can exceed the specified commitments and obligations.

• The AI convention does not introduce new or substantive human rights specific to AI. Instead, it emphasizes that existing human and fundamental rights, protected by international and national laws, must also be safeguarded in the application of AI systems.
• The convention primarily places obligations on governments, expecting them to implement effective remedies (Article 14) and procedural safeguards (Article 15).
• Overall, the convention adopts a comprehensive approach to mitigating risks that AI systems pose to human rights and democracy.
• Implementing the convention will undoubtedly face challenges, especially since AI regulation regimes are still developing and technology continues to advance faster than policy.

Global cooperation is essential. Countries must engage in international forums to harmonize AI regulations and standards, based on the AI convention. Regional frameworks should be developed. The convention can guide regions to create frameworks that address local challenges while aligning with core principles. Nationally, robust frameworks are crucial. Countries need to implement clear regulations and safeguards to protect human rights and democracy in the AI era.