U.S. Senate negotiators are considering legislation that would require artificial intelligence developers to design safer products to limit catastrophic risks, while also discussing whether to give the government authority to block the release of certain advanced models deemed unsafe.
People familiar with the matter said the proposed legislation would impose what is known as a “duty of care” on developers, requiring companies to design their products to prevent potential catastrophic risks.
Government Powers Under Negotiation
Negotiators are considering giving the government authority to ban the release of artificial intelligence models deemed unsafe, while allowing companies to challenge the bans in federal courts.
The mechanism for exercising these government powers remains under negotiation among the parties involved in the Senate.
Limiting State Laws
The proposal would also bar states from enforcing their own laws against certain risks associated with artificial intelligence models, including their potential use in designing nuclear or biological weapons.
