Advertisement

Commerce opens comments on regulating certain influential AI models

Dual-use foundation models, with open model weights, present particular risks.
(Getty Images)

The Commerce Department’s National Telecommunications and Information Administration is opening a request for comment on open foundation models, expressing particular interest in dual-use foundation models with “widely available weights.” The NTIA believes this particular AI technology raises critical safety and innovation questions because it can be extremely powerful — and theoretically used by anyone. 

The NTIA is interested in studying the risks that could be introduced when model weights for dual-use foundation models are released publicly. Eventually, the agency plans to release a report that could include policy recommendations. Importantly, the announcement comes as agencies continue to enact the Biden administration’s executive order on artificial intelligence, which the president signed last October. Many agencies are on track, while others are continuing to work on their assignments.

“These are some of the most important and consequential AI systems that are going to be developed,” Alan Davidson, assistant secretary of commerce for communications and information and NTIA administrator, said during a call with reporters. “Early conversations around AI openness and open foundation models have engendered fear about making the most advanced models widely available without adequate restrictions or safeguards against misuse.” 

The challenge is that while making this kind of AI generally available could accelerate innovation and help small businesses and startups, it could also introduce the risk of misuse. For instance, nefarious actors could remove safeguards built into AI models and deploy the technology in dangerous ways — and without the need for new training data. 

Advertisement

The request for comment, Davidson explained, is meant to study the risks of this approach to AI, look at different licensing and distribution models that could be available, and consider voluntary or mandatory regulatory models. 

Critically, senior administration officials acknowledged that NTIA does not have specific regulatory authority in this area. Officials also would not comment on which federal agencies might have a role in addressing this particular technology, while acknowledging the challenge of addressing dual-use foundation models with open weights based outside the United States. 

“AI is an accelerator — it has the potential to make people’s existing capabilities better, faster, and stronger. In the right hands, it carries incredible opportunity, but in the wrong hands, it can pose a threat to public safety,” Commerce Secretary Gina Raimondo said in a statement. This is an important piece of the president’s executive order and an early step toward ensuring safety, security, and trust in these systems.” 

In an emailed statement to FedScoop, Nick Clegg, Meta’s president for global affairs, said that the company looks forward “to working with the administration to share what we’ve learned from building AI technologies in an open way over the last decade so that the benefits of AI can continue to be shared by everyone.”

Comments to the NTIA are due within a month of the request’s publication in the Federal Register. 

Rebecca Heilweil

Written by Rebecca Heilweil

Rebecca Heilweil is an investigative reporter for FedScoop. She writes about the intersection of government, tech policy, and emerging technologies. Previously she was a reporter at Vox's tech site, Recode. She’s also written for Slate, Wired, the Wall Street Journal, and other publications. You can reach her at rebecca.heilweil@fedscoop.com. Message her if you’d like to chat on Signal.

Latest Podcasts