Maria Cantwell proposes federal safety framework for frontier AI

Maria Cantwell outlined a federal governance framework for frontier AI in a Senate Commerce Committee announcement. The proposal calls for enforceable safety standards, continuous testing and independent audits before covered models are released.

Under the framework, the National Institute of Standards and Technology would develop measurable, risk based standards for systems that could cause catastrophic harm. The standards would cover cyberattacks, chemical and biological threats, loss of human control, unsafe self improvement and autonomous agents escaping secure testing environments.

Developers would give qualified auditors access to evaluate models and would report serious safety and security incidents. The framework also calls for whistleblower protections, human review of consequential decisions, child safety measures, worker training and international coordination on AI standards and incident information.