Federation of American Scientists

Advanced model weights can be high-consequence assets. If stolen or openly released without adequate safeguards, they may allow actors to remove safety filters, fine-tune models for harmful uses, or reproduce capabilities that would otherwise remain ...

Support research into interpretability, robustness, control, and the preservation of monitorable AI reasoning.

Mandate or formalize AI model evaluation and disclosure for cyber risks. Establish requirements for structured, pre-deployment evaluations of frontier AI models focused on cyber-relevant risks, including misuse potential, autonomy, reliability over t...