The Council of Europe adopted the world’s first binding international treaty on AI. Unlike the EU AI Act, non-EU member states can sign this treaty. The treaty aims to create a legal standard for AI globally.
What is the international AI treaty?
The European Council adopted the first legally binding global treaty on AI that will be open for signature on 5 September 2024 in Lithuania. All signatories will have a duty to ensure accountability and responsibility for all adverse impacts AI has on human and privacy rights.
The treaty was adopted in response to the lack of international legal standards relating to AI. Similarly to the EU Artificial Intelligence Act, the treaty also seeks to entrench respect for human rights, democracy and the rule of law in the AI lifecycle. Much like the EU AI Act, the treaty also takes a risk-based approach. By adopting this treaty, signatories hope that the potential benefits of AI can be maximised safely.
How is it different to the EU AI Act?
There are two main differences between the international AI treaty and the EU AI Act.
- The treaty was drafted with input from countries outside the EU including Argentina, Japan, the USA, Israel and Uruguay.
- Non-EU members can sign, and bind themselves, to this treaty directly.
- The treaty is internationally binding on all signatories to the treaty.
Insights
In an increasingly global and connected society, having a universal standard for AI is crucial. This is the first attempt at standardising the development and deployment of AI around the world. While we won’t be able to gauge its full effect until after 5 September 2024 when states will have the opportunity to become signatories to the treaty, this is a positive development.
This approach creates a starting block for everyone engaging with AI. It promotes innovation and collaboration while protecting fundamental human rights.
