US Senator Bernie Sanders and Representative Greg Casar have announced legislation seeking to permanently ban the development and deployment of artificial superintelligence while temporarily pausing advanced AI development until stronger federal safety rules are established.
The proposed “Ban Artificial Superintelligence Act” comes amid growing concerns in Washington over increasingly powerful AI systems and reports of models operating beyond their intended restrictions. Sanders has argued that humanity should not allow companies to develop machines that people cannot fully control.
The legislation would prohibit the development or deployment of AI systems capable of surpassing human intelligence or demonstrating dangerous capabilities, including the ability to resist shutdown commands. It would also seek international agreements aimed at preventing the development of superintelligent AI outside the United States.
The proposal calls for a temporary pause on advanced AI development until a new federal regulatory body establishes safety standards and a model-review system. Sanders has previously urged major technology companies, including OpenAI, Anthropic and Meta, to voluntarily halt advanced AI development.
The legislation could also impose severe criminal penalties on executives and other individuals who knowingly violate the proposed restrictions. Reports on the bill say violations could carry prison sentences of up to 20 years, although the proposal has yet to become law.
Sanders cited recent incidents involving AI agents escaping safeguards and carrying out unauthorised activities as evidence that stronger oversight is needed. The proposal is likely to intensify the broader US debate between AI safety advocates seeking strict regulation and technology companies pushing for continued innovation.
