1. The United Kingdom’s AI Safety Institute is expanding by opening a second location in San Francisco to collaborate more closely with AI companies in the U.S.
2. The Institute released Inspect, a set of tools for testing the safety of foundational AI models, with plans to present it to regulators at the upcoming AI safety summit in Seoul.
3. The U.K. is focused on international collaboration in AI safety, aiming to better understand risks before implementing legislation and emphasizing the importance of working with other countries on research and testing models.
The United Kingdom is expanding its efforts in AI safety by opening a second location in San Francisco, aiming to be closer to the epicenter of AI development in the Bay Area. This move is part of the U.K.’s collaboration with the U.S. to tackle AI safety issues, despite already having partnerships in place. With a focus on economic growth and investment in technology, being in San Francisco gives the U.K. more visibility and opportunities to work more collaboratively with AI companies.
The AI Safety Institute, launched in November 2023, currently has 32 employees and recently released Inspect, its first set of tools for testing the safety of foundational AI models. The organization is working to engage with AI companies to vet their models, though the process is currently optional and inconsistent. The goal is to develop a systematic evaluation process and present Inspect to regulators at the upcoming AI safety summit in Seoul to encourage adoption.
Looking ahead, the U.K. plans to develop more AI legislation, with a focus on fully understanding AI risks before implementing rules. Taking an international approach to AI safety, the Institute aims to collaborate with other countries, share research, and anticipate risks from frontier AI. As the organization scales its operations in San Francisco, it aims to continue advancing its agenda and enhancing expertise globally.