What happened
On September 3, 2026, Sen. Bernie Sanders and Rep. Greg Casar announced the Ban Artificial Superintelligence Act, which would permanently ban development/deployment of AI systems that surpass human cognitive capability or could subvert shutdown/control, and impose a temporary pause on 'advanced AI' development until a new cabinet-level federal regulator establishes safety rules and a pre-deployment review process. Violations could carry up to 20 years imprisonment for individuals and forced corporate dissolution ('corporate death penalty'). The bill also calls for international agreements and export controls to pursue a global ban.
Why it matters
This is the most aggressive AI-restriction proposal introduced in Congress to date, reflecting escalating legislative concern following the OpenAI/Hugging Face agent breach and other autonomous-agent incidents. While passage is highly unlikely under the current administration (which just secured G20 backing for a deregulatory approach one day earlier), it marks a hardening of the congressional minority position and could shape future debate/amendments to other pending AI bills.
Action needed
Monitor for committee referral and co-sponsorship; frontier labs should track definitional scope of 'superintelligence' and 'advanced AI' as the bill develops, given potential future compliance exposure.