UK and US Forge Alliance on AI Safety Science, Paving the Way for Global Collaboration

Historic Partnership Seeks to Accelerate Development of AI Safety Testing Protocols.

In a landmark move, the United Kingdom and the United States have joined forces to advance the science of AI safety, signing a Memorandum of Understanding (MOU) aimed at developing rigorous tests for the most sophisticated artificial intelligence (AI) models. The collaboration, announced on Monday, April 1, 2024, marks a significant step towards fulfilling commitments made at the AI Safety Summit held last November.

The MOU, signed by UK Technology Secretary Michelle Donelan and US Commerce Secretary Gina Raimondo, outlines plans for both nations to harmonize their scientific approaches and collaborate closely in the rapid development of robust evaluation suites for AI models, systems, and agents.

Under the partnership, the UK and US AI Safety Institutes have unveiled plans to establish a unified approach to AI safety testing, with a focus on sharing capabilities and expertise to effectively address emerging risks. One of the key initiatives outlined in the agreement involves conducting joint testing exercises on publicly accessible AI models. Additionally, the Institutes aim to facilitate personnel exchanges to leverage a collective pool of expertise.

Effective immediately, the partnership aims to streamline collaboration between the two organizations, recognizing the urgent need to address the rapidly evolving landscape of AI technology. As AI continues to advance at a rapid pace, both governments emphasize the importance of adopting a shared approach to AI safety that can keep pace with emerging risks.