Sanders, Casar Propose US Superintelligence Ban With 20-Year Sentences

Artificial superintelligence refers to systems that would broadly exceed human cognitive abilities and potentially resist human shutdown or control, a narrower and more extreme category than today’s generative AI and the loosely defined concept of AGI. The issue has moved rapidly up Washington’s agenda after OpenAI and other frontier laboratories disclosed cases of agents breaching safeguards or acting beyond their assigned scope. The proposed ban marks a sharp shift from calls for testing and transparency toward an outright prohibition on the most capable systems.

On Sept. 3, 2026, Senator Bernie Sanders of Vermont and Representative Greg Casar of Texas announced the forthcoming Ban Artificial Superintelligence Act. It would permanently prohibit developing or deploying superintelligent AI, pause advanced AI work until a new federal regulator sets safety rules, expose individuals to as much as 20 years in prison and allow courts to dissolve offending companies. The same day, OpenAI released GPT-6 Astra, saying it could independently find unknown security flaws and develop new exploits, while committing $1 billion to expand its Daybreak cyber-defense initiative.