USA, UK and EU Sign World First AI Treaty

The world has taken a significant step toward aligning its goals and values concerning artificial intelligence (AI) following a meeting of Council of Europe justice ministers.

The United States, European Union, and the United Kingdom are expected to sign the Framework Convention on AI on September 5.

This treaty emphasizes the importance of human rights and democratic values in regulating AI models used by both public and private sectors.

The Framework Convention on AI would be the first legally binding international treaty on AI, holding signatories accountable for any harm or discrimination caused by AI systems.

It also requires that the outputs of these systems respect citizens’ rights to equality and privacy, providing legal recourse for those affected by AI-related violations.

However, penalties, such as fines, for violations have not yet been established.

At this stage, compliance with the treaty is enforced only through monitoring.

Peter Kyle, the UK’s minister for science, innovation, and technology, described the treaty’s signing as an "important" global first step.

“The fact that we hope such a diverse group of nations is going to sign up to this treaty shows that actually, we are rising as a global community to the challenges posed by AI,” Kyle remarked.

The treaty was initially drafted two years ago with input from over 50 countries, including Canada, Israel, Japan, and Australia.

While this treaty is the first international agreement on AI, individual nations have been working on their own AI regulations.

This past summer, the EU became the first jurisdiction to implement broad rules governing the development and deployment of AI models, particularly high-level ones with significant computing power.

The EU AI Act, which came into effect on August 1, introduced substantial AI regulations through phased implementation and key compliance requirements.

Though the legislation was designed with safety in mind, it has drawn criticism from AI developers who argue that it hinders innovation within the region.

Meta, the parent company of Facebook, halted the rollout of its latest AI products, including Llama2, in the EU due to regulations that restrict access to the most advanced AI tools for Europeans.

In August, several tech firms wrote a letter to EU leaders requesting more time to comply with the new regulations.

In the United States, Congress has yet to introduce a national framework for AI regulation, but the Biden administration has established AI safety committees and task forces.

Meanwhile, California has been actively drafting and passing AI regulations. Recently, two bills passed the State Assembly and await a decision from Governor Gavin Newsom.

One bill regulates and penalizes the creation of unauthorized AI-generated digital replicas of deceased personalities.

The second, a controversial bill opposed by many top AI developers, mandates safety testing for most advanced AI models and requires provisions for a "kill switch" for such models.

California’s AI regulations are particularly significant, as the state is home to major developers like OpenAI, Meta, and Alphabet.

Sources:

https://cointelegraph.com/news/global-ai-treaty-human-rights-regulation

https://www.ft.com/content/4052e7fe-7b8a-4c42-baa2-b608ba858df5