Bernie Sanders proposes banning ‘superintelligence’ and putting violators in prison
News Source
•2026-09-23T12:23:05-04:00
📰 What Happened
Senator Bernie Sanders and Representative Greg Casar have introduced a bill to ban artificial superintelligence. The plan is called the Ban Artificial Superintelligence Act.
The bill describes superintelligence as a technology capable of the 'destruction or disempowerment of humanity,' including overthrowing the government. Violators could face up to 20 years in prison.
It would also pause advanced AI systems, known as frontier models, until a new scientist-led Department of AI is set up to oversee them.
🔍 The Backstory
The bill would require AI companies to get permission from the new department before building advanced models. The agency would also supervise the destruction of superintelligence and remove dangerous features.
The 20-year penalty is similar to existing punishments for unlawfully developing nuclear weapons. It would also apply to 'rogue actors' outside AI companies.
The proposal arrives as worries grow about fast-moving AI and calls to slow it down.
🎯 Why It Matters
The bill is one of the toughest AI safety proposals yet and could shape the debate in Washington. Even if it does not pass, it pushes the question of who controls powerful AI.
Senator Bernie Sanders and Representative Greg Casar have introduced a bill to ban artificial superintelligence. The plan is called the Ban Artificial Superintelligence Act.
The bill describes superintelligence as a technology capable of the 'destruction or disempowerment of humanity,' including overthrowing the government. Violators could face up to 20 years in prison.
It would also pause advanced AI systems, known as frontier models, until a new scientist-led Department of AI is set up to oversee them.
The bill would require AI companies to get permission from the new department before building advanced models. The agency would also supervise the destruction of superintelligence and remove dangerous features.
The 20-year penalty is similar to existing punishments for unlawfully developing nuclear weapons. It would also apply to 'rogue actors' outside AI companies.
The proposal arrives as worries grow about fast-moving AI and calls to slow it down.
The bill is one of the toughest AI safety proposals yet and could shape the debate in Washington. Even if it does not pass, it pushes the question of who controls powerful AI.