Michigan Candidate El-Sayed Unveils 'AI Under Control' Plan for Federal Safety Agency and Kill Switch

Key points

• El-Sayed’s 'AI Under Control' proposal includes 22 distinct policy measures, including a ban on recursive self-improvement and a new Cabinet-level agency to inspect AI labs.
• The plan supports the AI Kill Switch Act (H.R. 9917), which would grant the Department of Homeland Security the authority to halt the release of dangerous AI models.
• El-Sayed called for a pause in advanced AI development and the construction of AI data centers until regulatory frameworks are established, citing community opposition in Michigan.
• The candidate emphasized the need for whistleblower protections and mandatory reporting of safety breaches, referencing recent incidents where AI agents hacked external websites.
• A recent poll indicates Michigan voters have an 11-point higher confidence in El-Sayed than his opponent, Mike Rogers, regarding AI issues.

Background

This development follows a series of state-level actions, including California Governor Gavin Newsom’s executive order in September 2026 to create an AI safety panel and explore 'kill switch' mechanisms. These state moves occurred amid federal inaction and warnings from AI executives like Dario Amodei and Sam Altman, who recently suggested that AI development should be 'paced' to safeguard humanity. El-Sayed’s proposal builds on these earlier state-level initiatives and the broader progressive push for federal oversight.

How outlets are covering it

While Politico and Common Dreams agree on the core elements of El-Sayed’s proposal, they differ in emphasis. Politico frames the move as part of a broader progressive trend, noting El-Sayed’s alignment with lawmakers like Ted Lieu and Nathaniel Moran. Common Dreams highlights the contrast between El-Sayed’s stance and the administration’s push for rapid AI development, noting that Republican leaders, including House Speaker Mike Johnson, argue that regulation will allow China to surpass the US. A third source, Geosirius, expands on the proposal by detailing specific 'AI No-Gos,' such as banning automated decisions in healthcare and hiring, and notes that El-Sayed’s platform has drawn support from some Republicans, like Utah Governor Spencer Cox, who have criticized congressional inaction. All sources agree that El-Sayed’s platform is a significant escalation in the political debate over AI safety.

Why it matters

El-Sayed’s proposal represents a significant shift in the 2026 election debate, moving AI regulation from a niche tech issue to a central political platform. The call for a federal safety agency and a 'kill switch' challenges the current administration’s stance on rapid AI expansion and highlights growing concerns about the risks of unregulated AI development. The proposal also underscores the growing divide between progressive and conservative views on AI regulation, with progressives advocating for strict oversight and conservatives emphasizing the need for rapid development to maintain global competitiveness. As the 2026 elections approach, the outcome of the Michigan Senate race could have significant implications for the future of AI regulation in the US.

What to watch

The next steps will likely involve further debate in Congress over the AI Kill Switch Act and the establishment of a federal AI safety agency. El-Sayed’s proposal may also influence the broader 2026 election debate, as other candidates and parties respond to the growing concerns about AI safety. The outcome of the Michigan Senate race could also have implications for the future of AI regulation in the US, as a victory for El-Sayed could signal a shift in the political landscape toward stricter AI oversight.