AI researcher Gary Marcus published a critique of the Ban Artificial Superintelligence Act introduced by Senator Bernie Sanders and Representative Greg Casar on September 3, 2026. The bill would permanently ban development of AI systems matching human cognitive abilities and impose a temporary pause on advanced AI development, with penalties including up to 20 years in prison and a “corporate death penalty” for violations. Marcus argues that while he supports a temporary pause and a new AI regulatory agency, a permanent, unilateral ban is overly broad, could put the U.S. at a competitive disadvantage, and rests on an unclear technical definition of the superintelligence threshold. He proposes an alternative that would permit narrow AI tools while requiring developers to demonstrate alignment, control, and oversight to independent authorities before deploying more powerful systems.
