On 15 March 2025, the National Institute of Standards and Technology (NIST) closes the public consultation on the updated guidelines on managing misuse risk for dual-use foundation models. The guidelines offer voluntary best practices to identify and mitigate risks to public safety and national security in the Artificial Intelligence (AI) lifecycle. It provides a framework for mapping, measuring, managing, and governing risks while promoting transparency and collaboration. The updated guidelines also include a new appendix on model evaluations, expanded guidelines on cybersecurity and chemical/biological risks, and a "marginal risk" framework. It also contains updates to support open model developers and additional content for managing risk across the AI supply chain.
Original source