TechFlow news: OpenAI and Anthropic have reached an agreement allowing the U.S. government to access their new AI models before and after release to enhance safety.
The U.S. AI Safety Institute announced on Thursday that this move will help assess security risks and mitigate potential issues. California recently passed the "Safe and Secure Innovation for Frontier Artificial Intelligence Models Act" (SB 1047), requiring AI companies to implement safety measures, sparking industry opposition over concerns it could impact small open-source developers. The White House is also seeking voluntary commitments from major companies to advance AI safety initiatives. Elizabeth Kelly, director of the U.S. AI Safety Institute, said the new agreement marks a significant milestone in responsibly governing the future of AI.
