It would provide “clarity” to allies about AI’s benefits, the White House said.
The Biden administration is rolling out a new rule that is aimed at responsible use of artificial intelligence technology.
According to a press release about the rule, this action “streamlines licensing hurdles” and “provides clarity to allied and partner nations about how they can benefit from AI.”
The action “is designed to safeguard the most advanced AI technology and ensure that it stays out of the hands of our foreign adversaries, but also enabling the broad diffusion and sharing of the benefits with partner countries,” Commerce Secretary Gina Raimondo said during a call with reporters about the move.
“The focus is on the frontier, the most advanced AI models and the largest compute clusters,” she added.
The rule has a three-pronged approach, Raimondo said: “expanding and updating controls for advanced AI chips,” “creating a new set of controls for the most advanced, closed AI model weights make sure they don’t fall into the hands of our adversaries” and “imposing security conditions to safeguard critical technology and the largest AI clusters.”
She noted that some industries would not be impacted because they are not crucial to national security, including supply chain activities and gaming chips.
Raimondo said that the rule is very complex and, because of that, the comment period is 120 days, longer than usual timeline for rule-making. She also noted that this rule comes with just about one week left of President Joe Biden’s presidency.
“I fully expect the next administration may make changes as a result of that input,” Raimondo said. “So we we’ve provided for 120 days, which is very long comment period, and we provided for one year, 365 days for compliance for the standards at AI data centers to make sure industry is fully cited on the new rules and able to comply.”
But one official on the call said this effort has bipartisan support, especially because this is something that concerns national security.
National Security Adviser Jake Sullivan said that this rule has been in development for a long time with key shareholders.
“There are many leading AI developers who are projecting that AI capabilities will exceed human capabilities in fields from physics to biology to electrical engineering in the very near future, and that has economic and technological implications,” Sullivan told reporters, “but it also has fairly profound national security implications as well. So from our perspective, we have a national security responsibility to do two things, first, to preserve, protect and extend American AI leadership, particularly vis a vis strategic competitors.”
