Sanders and Casar propose ban on artificial superintelligence

Sen. Bernie Sanders and Rep. Greg Casar introduced legislation Wednesday that would ban the development of artificial superintelligence and pause work on other advanced AI systems until the federal government establishes safety oversight.

The 19-page Ban Artificial Superintelligence Act would create a Cabinet-level Department of Artificial Intelligence, led by a secretary of AI. The department would regulate the industry and supervise the most advanced models. Under the bill, it could remove dangerous models from circulation and oversee their destruction, Al Jazeera reported.

The proposed ban targets AI that could outperform humans and has the power “to destroy or disempower humanity, including by overthrowing the federal government,” according to Al Jazeera. NBC News reported that the bill also lists the capacity to automate or greatly accelerate AI research and development among the warning signs it seeks to address. An aide involved in drafting the measure told NBC News that the goal is to prevent AI systems from improving themselves in ways that could escape human control.

Violations could bring up to 20 years in prison, a penalty Sanders compared with punishment for unlawfully building nuclear weapons. Al Jazeera reported that individuals who violate the provisions would also be barred from working in the AI industry for 10 years. Companies could face what the bill’s backers call a “corporate death penalty,” including federal seizure of their assets and intellectual property, the outlet reported.

The measure also calls for international agreements to establish a global ban and export controls on AI computing infrastructure, according to Al Jazeera.

“We’re dealing with something that humanity has never seen before,” Sanders told reporters Wednesday, calling AI “the most transformational technology in the history of the world.” He said Congress was doing “virtually nothing” to prepare for its effects.

The proposal drew support from ControlAI, an advocacy group seeking restrictions on the technology. Its leaders praised the bill for defining superintelligent AI partly by its capacity to overpower the federal government. Kevin Frazier, an AI policy expert at the University of Texas School of Law, told NBC News that Congress needs to act quickly, though he said the bill overlooks other ways to assess companies’ safety practices.

Critics questioned both its scope and its prospects. Alex Turner, a former Google DeepMind researcher, told Al Jazeera the bill was “unlikely to pass” and said one of its definitions might cover current systems. Caleb Max, head of the National AI Association trade group, told NBC News the measure would be “dead on arrival with both the AI industry and Washington.” He also questioned whether its penalties were constitutional.

Other senators are pursuing different approaches. Sen. Maria Cantwell told NBC News federal rules should include standards, independent testing and international incident reporting. Sen. Ted Cruz said he was negotiating a separate bill with Sens. John Thune and Amy Klobuchar to address catastrophic risks, but they had not reached an agreement. NBC News reported that a proposal from Sen. John Kennedy requiring AI companies to have an emergency shutoff was blocked by Sen. Rand Paul.

The debate comes as President Donald Trump said Tuesday that government documents would use “super intelligence” instead of “artificial intelligence.” With Congress nearing an election recess, NBC News reported there is little expectation that major AI legislation will pass this year.