Senate Commerce Committee ranking member Maria Cantwell unveiled a new AI safety framework that assigns a central role to the federal government in setting safety standards and mandating rigorous testing of models that could pose catastrophic risks. The proposal, released Wednesday, reflects growing bipartisan concern over unchecked AI development.
Cantwell’s plan would create a federal testing regime overseen by the Commerce Department, requiring developers to submit model architectures, training data, and safety audits before deployment. By establishing uniform benchmarks, the framework aims to prevent scenarios where powerful AI systems escape human control, a goal that aligns with executive orders on AI risk management.
Industry observers say the proposal could become a cornerstone for future legislation, but warn that prescriptive testing may stifle innovation. Cantwell’s office has begun consultations with tech leaders, signaling a collaborative approach. If adopted, the framework could set a template for balancing AI’s promise with public safety.
Source: Read original article