Democrats Considering Crackdown That Would Require Government Permission to Train AI Models

The incoming Democratic majority in Congress appears poised to place stringent controls on the rapidly advancing AI industry. In addition to regulations concerning data center energy consumption and product liability, lawmakers are reportedly contemplating new measures to rein in the sector before its unchecked growth leads to potential harm.

Arizona Democrat Yassamin Ansari called for a "pause on advanced AI development, regulation of this out of control industry, and a ban on artificial superintelligence" as the Democratic agenda for the upcoming legislative session. House Minority Leader Ted Lieu echoed similar sentiments, chastising Republicans for allowing the AI sector to "run wild" following the recent election.

A bill spearheaded by representatives Greg Casar, Alexandria Ocasio-Cortez, and Senator Bernie Sanders, dubbed the Ban Superintelligence Act, is gaining traction. This legislation would establish a permitting process administered by a new Department of Artificial Intelligence, granting the agency authority to issue credentials to AI labs working on advanced AI systems.

The draft bill mandates that AI labs grant total access to their digital and physical infrastructure, encompassing data centers, to the AI secretary, who would have the power to define "advanced" AI through a fluid threshold adjusted according to technological advancements. The bill also outlines severe penalties for non-compliant entities, including 20-year prison sentences for the gravest offenders and a 10-year ban on industry employees engaging in such activities again.

While the extent of support from more centrist Democrats remains uncertain, passage of the Ban Superintelligence Act would undoubtedly have far-reaching consequences for an AI industry that has thus far operated under minimal oversight in the United States.