Lawmakers push for mandatory AI safety standards and federal transparency
Democratic Reps. Sara Jacobs and Don Beyer are drafting a bill to mandate minimum safety standards for AI labs and grant the government emergency authority to shut down models that fail testing. Simultaneously, Sen. Gary Peters introduced the Transparent Automated Governance Act to require federal agencies to disclose AI use in critical decisions. These moves contrast with the administration's reliance on voluntary industry standards.
Key points
- Jacobs and Beyer's proposed framework includes liability clarifications and mandates for federal oversight agency resources.
- The bill was previewed at an invite-only conference in Berkeley, California, last weekend.
- Sen. Gary Peters' TAG Act requires federal agencies to notify individuals when AI is used in critical decisions and provide human review for adverse outcomes.
- Sen. Patty Murray criticized the lack of federal regulation, urging uniform national laws over state-by-state measures.
- The Trump administration has favored voluntary self-regulation, recently hosting a summit where major tech companies agreed to safety standards.
Background
This legislative activity follows a September debate shifting from existential AI fears to concrete regulatory frameworks. It also occurs after Sen. Maria Cantwell released a six-point safety framework, positioning her as a potential leader on AI policy if Democrats regain Senate control in the 2026 midterms. Earlier, OpenAI unveiled a framework for disclosing model misalignment incidents, and the US and China agreed to an AI dialogue mechanism.
How outlets are covering it
Politico highlights the proactive stance of Jacobs and Beyer, emphasizing mandatory safety standards and emergency shutdown powers. The Alpena News focuses on Peters' bipartisan effort to ensure transparency in federal AI usage, framing it as a consumer protection measure. The Spokesman-Review contrasts these legislative efforts with the administration's 'hands-off' approach, noting Murray's criticism of voluntary standards as insufficient. While all sources agree on the need for regulation, they differ in focus: Politico and the Spokesman-Review emphasize safety and risk mitigation, while the Alpena News emphasizes transparency and individual rights in government interactions.
Why it matters
These proposals represent a significant shift from voluntary industry commitments to mandatory federal oversight. If enacted, they could establish the first comprehensive legal framework for AI safety and liability in the US, potentially influencing global standards and addressing public concerns about AI risks without waiting for catastrophic failures.
What to watch
The next steps include the formal introduction of the Jacobs-Beyer bill and the TAG Act in Congress. Sen. Cantwell's framework may also gain traction if Democrats win the Senate in the 2026 midterms. The administration's response to these legislative efforts will determine whether federal AI regulation moves forward or remains voluntary.
- AI safety framework floated by Democratic duo Politico
- Cantwell unveils AI safety framework as Congress stalls ahead of election The Seattle Times
- Peters introduces bipartisan bill to require transparency about federal government’s use of AI The Alpena News
- Reps. Carter, Thompson Reintroduce the Cybersecurity and Infrastructure Security Agency Securing AI Task Force Act to Address the Evolving Challenges Posed by AI House Homeland Democrats (.gov)
- Sen. Patty Murray meets with state policymakers to discuss AI regulations as federal action sputters The Spokesman-Review
Want the full story? Read the original reporting
Read on Politico