Sanders and Casar introduce bill to ban superintelligence development

The bill's most consequential detail is definitional: by codifying 'superintelligence' as a distinct legal category subject to permanent prohibition rather than conditional oversight, Congress would be setting a statutory ceiling on capability development, not just a procedural speed bump. That framing has no real precedent in U.S. technology law.

The timing sits in sharp tension with what else is happening in the space right now. Lovable's $600M annualized revenue milestone, covered here the same day, illustrates how rapidly AI-assisted development is becoming consumer-scale infrastructure. A federal agency tasked with enforcing a capability ceiling would need to define where 'useful AI tooling' ends and 'prohibited superintelligence' begins, and the commercial momentum documented in that Lovable story makes that boundary harder, not easier, to draw. The broader legislative context is largely disconnected from recent funding and product coverage in the archive, placing this story more squarely in the emerging governance beat than the commercial AI beat.

Watch whether the bill attracts co-sponsors from outside the progressive caucus within the next 60 days. Bipartisan support would signal genuine legislative traction; absence of it would mark this as a pressure document rather than a viable enforcement framework.