US Senate negotiators consider requiring AI firms to mitigate known major risks - Reuters

The account is attributed to Senate aides and a lobbyist. The report describes the proposal as under debate among negotiators rather than as a measure that has been introduced or passed.

The report says the requirement on companies would be to design AI products to prevent catastrophic risks, and that the government would hold the power to block the release of certain models it deems unsafe. Under the arrangement described, a decision to block a model's release would bring the federal courts into the process.

The article places the proposal within a Senate negotiation over what obligations AI developers would carry and what authority the federal government would hold over their products. It does not name the senators involved, the specific models that might be affected, the agency that would make a determination of unsafety, or the definition of the 'catastrophic risks' the design requirement targets. The excerpt does not state a timeline for the negotiations or how the measure would relate to existing law.