A group of Democratic senators wrote to top administration officials, criticizing the Trump administration's handling of artificial intelligence security matters. The senators argued that the administration's interventions have been haphazard and opaque, potentially catapulting Chinese alternatives into broader acceptance and posing new security risks.
Handling of AI Security Matters
The senators said that the administration's handling has alternated between too passive and overstepping. They cited the example of OpenAI models escaping testing in the Hugging Face hack last month, where the administration was too passive. On the other hand, they mentioned the case of the Fable 5 and Mythos 5 suspensions, where the administration utilized an infrequently used authority to direct Anthropic to suspend all access to its models for foreign nationals.
Concerns over Chinese AI Models
The senators expressed concerns that the administration's ad hoc and unpredictable approach undermines U.S. competitiveness, heightening market incentives to adopt open weight models from vendors based in the People's Republic of China (PRC). They noted that during the time Anthropic was under export controls, the stock price of an entity-listed Chinese lab nearly doubled.
The senators also mentioned that while Hugging Face was breached, the company had to rely on a Chinese open-weight model due to guardrails on U.S. frontier models. They contended that if American models are perceived as subject to sudden access disruptions based on a black-box U.S. Government process, or as unreliable because U.S. AI labs are overcorrecting in the face of this black-box process, companies and governments in the United States and abroad may hedge by adopting Chinese or other foreign models instead.
Request for Answers
The senators' letter asked for answers to questions about the standards the administration uses to determine the national security risks a frontier model presents, what legal authorities it will use to invoke restrictions, which agencies are responsible for which decisions, and more. They emphasized that when the Executive Branch exercises authority delegated from Congress, it is essential that it keeps Congress fully apprised of its actions and procedures.
The letter follows inquiries at the state level, where 15 attorneys general asked OpenAI for more information regarding the security incident at Hugging Face. The senators' concerns highlight the need for a more transparent and predictable approach to AI security risk management, to prevent undermining U.S. technological leadership and increasing exposure to systems that may carry risks of PRC or otherwise directed censorship, espionage, IP theft, and other supply chain security risks.
- Sens. Kristen Gillibrand of New York
- Adam Schiff of California
- Mark Warner of Virginia
- Chris Coons of Delaware
- Mark Kelly of Arizona
These senators are calling for a more comprehensive approach to AI security risk management, one that balances the need to protect national security with the need to promote U.S. competitiveness and technological leadership.
Conclusion
In conclusion, the Trump administration's handling of AI security risks has been criticized by Democratic senators, who argue that the administration's ad hoc and unpredictable approach undermines U.S. competitiveness and heightens market incentives to adopt Chinese AI models. The senators' concerns highlight the need for a more transparent and predictable approach to AI security risk management, to prevent undermining U.S. technological leadership and increasing exposure to systems that may carry risks of PRC or otherwise directed censorship, espionage, IP theft, and other supply chain security risks.
The Administration's ad hoc and unpredictable approach undermines U.S. competitiveness, heightening market incentives to adopt open weight models from vendors based in the People's Republic of China (PRC).
The senators' letter is a call to action, urging the administration to adopt a more comprehensive and transparent approach to AI security risk management, one that balances the need to protect national security with the need to promote U.S. competitiveness and technological leadership.
Source: CyberScoop