What happened
On September 3, 2026, Sen. Bernie Sanders (I-VT) and Rep. Greg Casar (D-TX) announced the forthcoming Ban Artificial Superintelligence Act, which would permanently ban development/deployment of superintelligent AI systems (defined to include systems that surpass human intelligence generally or can subvert shutdown commands), temporarily pause advanced AI development until a new federal AI regulatory body establishes safety rules and a model review process, establish a new cabinet-level federal agency (advised by an AI Advisory Board) to enforce the ban, and direct the US to pursue international agreements against superintelligence development globally.
Why it matters
This is the most aggressive federal legislative AI-safety proposal introduced to date, calling for a moratorium on frontier development and a new cabinet-level regulator — a direct counterpoint to the same week's G20 'Carolina Principles' push for light-touch, no-new-regulator AI governance. It signals a widening partisan and institutional split in US AI policy.
Action needed
Frontier AI labs and deployers should monitor legislative momentum; while passage is unlikely near-term, the bill anchors the aggressive end of the US AI regulatory debate and could influence state-level or agency rulemaking framing.