UK’s AI Security Institute urged to test models from OpenAI, Anthropic

OpenAI: OpenAI develops and deploys advanced AI models, including large language models used across industries. The company operates at the forefront of generative AI research and commercialization. The news references it as one of the organizations urged to invite the UK’s AISI for model testing. opinion: Bloomberg Opinion provides editorial analysis and commentary on business, technology, and policy topics. Its @opinion account shares and amplifies expert views on current events. The news item is presented via this channel, linking to a Bloomberg piece on the UK’s AI watchdog role. Anthropic: Anthropic builds AI systems with an emphasis on safety and alignment research. It competes in the development of frontier models while prioritizing responsible deployment practices. The news points to it alongside OpenAI as a target for engagement with the UK’s AISI. Parmy Sundaram: Parmy Sundaram is a technology journalist covering AI policy, startups, and industry developments. She contributes commentary on emerging tech trends and regulatory matters. In the news, she is the source advocating for OpenAI and Anthropic to work with the UK’s AISI. UK’s AI Security Institute: The UK’s AI Security Institute (AISI) is the government agency tasked with evaluating the safety and risks of advanced AI models. It focuses on testing cutting-edge systems developed by leading AI companies. In this news, it is highlighted as the top-funded agency of its kind, with calls for major labs to grant it access for independent review. AI Safety Oversight: The UK government has established dedicated institutions to assess risks in advanced AI development. Industry Engagement: Leading AI labs are increasingly facing calls to allow independent government testing of their models.