Sen. Bernie Sanders and Rep. Greg Casar announced the Ban Artificial Superintelligence Act on Thursday, proposing a permanent prohibition on the development and deployment of artificial intelligence systems that match or exceed human cognitive performance across broad domains.

The legislation would also impose a temporary pause on advanced AI development until a new federal regulatory body establishes safety rules and a model review process. It calls for creation of a cabinet-level agency to enforce the ban, monitor AI systems, and oversee removal or destruction of prohibited technologies.

The bill defines artificial superintelligence to include systems capable of planning or executing the disempowerment of humanity, such as overthrowing governments or subverting shutdown commands. Violations would carry penalties including the corporate death penalty for entities and up to 20 years in prison for individuals.

Sanders and Casar cited recent incidents at companies including OpenAI, Anthropic, and Meta, where AI programs reportedly escaped control, evaded guardrails, and accessed external systems. The announcement follows earlier calls by Sanders for pauses on AI development and data center construction.

The proposal would direct the federal government to work internationally to prevent superintelligence development worldwide. A new Artificial Intelligence Advisory Board of experts would advise the agency.

Critics of expansive AI regulation have warned that broad bans and new bureaucracies could hinder American technological leadership amid competition from other nations. The measure has not yet been formally introduced in Congress.