What happened
On 24 September 2026, Rep. Tom Kean Jr. (NJ-07) introduced the AI Emergency Button Act (H.R.10567), which would require developers of AI systems to build in a technical capability for a human operator to shut the system down. Sen. John Kennedy (R-LA) introduced the Senate companion earlier in September. The bill is narrower than the earlier Lieu/Moran kill-switch proposal, applying only to the shutdown requirement.
Why it matters
It would convert the widely debated 'kill switch' from a voluntary safety best practice into a statutory mandate on every AI system, following the week's OpenAI agent incidents — part of the wave of in-window congressional bills responding to frontier-model concerns.
Action needed
AI developers building advanced systems should track this bill and the Senate companion; if momentum builds it would impose a hard engineering requirement (human-controlled emergency shutdown) on all AI systems rather than a voluntary 'kill switch' best practice.