Bernie Sanders proposes banning ‘superintelligence’ and putting violators in prison

Senator Bernie Sanders and Representative Greg Casar have introduced groundbreaking legislation aimed at prohibiting the development of artificial superintelligence, marking a significant shift in how lawmakers approach advanced AI governance. The proposed bill would classify superintelligence—defined as AI systems capable of causing "destruction or disempowerment of humanity"—as an illegal technology, with severe criminal penalties for violators. This legislative proposal reflects growing concerns among policymakers about existential risks posed by increasingly powerful artificial intelligence systems.

The Sanders-Casar bill represents one of the most aggressive regulatory approaches to AI development introduced in Congress. The legislation would establish criminal penalties for individuals and organizations attempting to develop superintelligent AI systems, though specific imprisonment terms remain part of the broader framework designed to deter such development. The bill defines superintelligence broadly as any AI system capable of governmental overthrow or large-scale human disempowerment, emphasizing the existential nature of the perceived threat.

The proposed framework demonstrates lawmakers' intent to move beyond advisory guidelines toward enforceable legal restrictions on AI advancement in particular directions.

• Criminal liability creates unprecedented legal risk for AI researchers and companies pursuing advanced AI development
• The definition of "superintelligence" may prove challenging to standardize and enforce across international boundaries
• Legislation could reshape investment priorities in the AI sector, redirecting funding toward regulated applications
• International cooperation becomes critical, as development could shift to jurisdictions without similar restrictions
• Establishes precedent for treating AI development as a criminal matter rather than purely regulatory concern
• May incentivize more transparent reporting of AI capabilities among major technology companies

The Sanders-Casar proposal signals that existential AI risks have entered mainstream political discourse at the legislative level. Rather than focusing solely on bias, privacy, or labor concerns, this approach treats superintelligence as a categorical threat requiring criminal prohibition. While the bill faces significant technical and enforcement challenges, it reflects a fundamental shift in how policymakers conceptualize AI governance—moving from industry self-regulation toward criminalized restrictions on specific technological pursuits. As AI capabilities advance, this legislative framework may influence how other democracies approach similar existential risk mitigation strategies.