The White House announced Monday it met a deadline to establish a voluntary framework for evaluating advanced AI models. The administration has not publicly released the specific details of the framework.
The strategy addresses national security and societal risks while avoiding regulations that could stifle innovation. This initiative follows a June 2nd executive order tasking several agencies with its creation.
AI developers may voluntarily submit their most advanced models for a pre-release review of up to 30 days. This process assesses potential security risks and is expected to influence industry standards.
The administration developed the framework through consultations with major AI labs, including Google, OpenAI, and Anthropic.