The United States and the United Kingdom signed a memorandum of understanding to work together to develop tests for more advanced AI models, fulfilling commitments made at the AI Security Summit last November, a statement said. From the US Department of Commerce.
The agreement, signed by US Commerce Secretary Gina Raimondo and UK Technology Minister Michelle Donnellan, will allow both countries to align their scientific approaches and work closely to accelerate and replicate robust sets of evaluations of models, AI systems and agents.
AI Safety Institutes in the US and UK have outlined plans to build a joint approach to AI safety testing and share their capabilities to ensure these risks can be effectively addressed. They intend to conduct at least one joint testing exercise in a publicly available form. They also plan to leverage the collective pool of expertise by exploring staff exchanges between institutes.
The partnership will take effect immediately and aims to allow both organizations to work seamlessly with each other. AI continues to evolve rapidly, and both governments recognize the need to act now to ensure a common approach to AI safety that can keep pace with emerging risks in the technology. As countries strengthen their AI safety partnership, they have also committed to developing similar partnerships with other countries to advance safety around the world.
“AI is the defining technology of our generation. This partnership will accelerate the work of both institutes across a broad range of risks, both to our national security and our broader society. Our association makes clear that we do not run away from these concerns, but rather confront them head-on. Through our collaboration, our institutes will gain It will better understand AI systems, conduct more robust evaluations, and issue more stringent guidelines.
“Subtly charming bacon junkie. Infuriatingly humble beer trailblazer. Introvert. Evil reader. Hipster-friendly creator.”