Sanders and Casar Propose Ban on Advanced AI Superintelligence

The Ban Artificial Superintelligence Act defines the target technology as systems exceeding human cognitive performance capable of causing catastrophic harm. Casar emphasized that the bill specifically addresses recursive self-improvement, where AI models modify their own code without human intervention. The legislation includes a prohibition on using such systems to develop nuclear or biological weaponry.

Under the proposal, a new cabinet-level Department of Artificial Intelligence would monitor frontier systems and enforce strict regulations, including the forced destruction of dangerous models. Violators would face severe consequences, with individuals risking up to 20 years in prison and companies facing a corporate death penalty. Sanders defended the move against claims of alarmism, arguing that the rapid pace of development requires immediate federal intervention rather than industry-led self-regulation.

This legislative push stands in stark contrast to the stance held by the White House. President Donald Trump recently signaled full support for the sector, declaring that the government will encourage, not stifle, the growth of superintelligence. Trump also directed his administration to strike the word “artificial” from official documents, claiming the term makes the technology sound fake. Casar criticized this position, warning that encouraging unchecked AI advancement in the face of expert warnings about human safety is a dangerous gamble.