President Joe Biden’s administration announced a new rule aimed at making it easier to sell artificial intelligence systems and computing power to U.S. allies, while keeping the technology out of the hands of countries of concern.
The interim final rule on the spread of artificial intelligence builds on previous regulations such as the chip controls announced in October 2022 and October 2023, the White House said Monday (Jan. 13). press release.
“To strengthen U.S. national security and economic strength, it is essential that we do not offshore this critical technology and that global AI runs on American rails,” the statement said. “It is important to work with AI companies and foreign governments to establish essential standards of security and trust as they develop their AI ecosystems. »
To facilitate the responsible diffusion of AI, the rule places no restrictions on chip sales to 18 U.S. allies and partners and streamlines the processing of low-risk shipments of the technology, according to the release.
The rule also creates two types of statuses – Universal Verified End User (UVEU) and National Verified End User – which help responsible and trusted entities benefit from U.S. technology, the release said.
It also allows the purchase of large amounts of computing power by non-VEU entities located outside of major allies and allows governments to double their chip cap by signing government-to-government agreements that align their values related to technology. ‘AI on those of the United States. according to the press release.
To restrict countries of concern’s access to technology, the rule contains provisions that ensure they can access advanced semiconductors for general applications but not for training advanced AI systems, restrict weight transfer of models and set safety standards to protect the weights of the countries concerned. advanced closed-weight AI models, according to the release.
Biden issued the first national security memorandum on AI in October, directing federal agencies to protect the country. AI advances as strategic assets while promoting their safe development for national security.
The White House directive sets out three main goals: maintaining U.S. leadership in the safe development of AI; harnessing AI for national security while protecting democratic values; and build international consensus on AI governance.
For all PYMNTS AI coverage, subscribe daily AI Newsletter.
