Washington DC [US], May 28 : Chairman John Moolenaar and Ranking Member Raja Krishnamoorthi of the House Select Committee on the Strategic Competition Between the United States and the Chinese Communist Party (CCP) are advocating for the Department of Commerce to broaden the responsibilities of the US AI Safety Institute (AISI) to directly tackle the increasing national security challenges posed by China's advancements in artificial intelligence, as stated in a release from the Select Committee on the Chinese Communist Party (SCCCP).
In a letter to Commerce Secretary Howard Lutnick, the lawmakers stressed that "there is a strong national security need for better understanding, predicting, and preparing for the PRC's AI progress," according to the SCCCP's statement.
They cited the January 2025 rollout of DeepSeek's large language model, R1, as a significant alert. The Committee's investigation into DeepSeek revealed "multiple national security risks, including the funnelling of Americans' private data to the PRC, manipulation of the model's outputs under PRC law, and the potential theft of US AI technology through model distillation," according to the SCCCP release.
As AI systems advance, the lawmakers cautioned that "the need to predict and prepare for PRC AI developments and to avoid strategic surprise will only grow more urgent," as Chairman Moolenaar and Ranking Member Krishnamoorthi noted in the SCCCP release.
The Committee urged AISI to expand its focus to specifically address the development of PRC AI, indicating that "a whole-of-government approach will be needed to ensure enduring US dominance in frontier AI development." They also pointed out AISI's "unique technical expertise, strong partnerships with industry, and experience in testing and evaluations" as valuable assets in the United States' efforts to outpace the CCP.
Moolenaar and Krishnamoorthi specified particular areas where AISI could bolster US national security, including evaluating the national security capabilities and vulnerabilities of the most sophisticated PRC AI models. AISI could conduct benchmarking and assess PRC models against US models to evaluate the competitive landscape, according to the SCCCP release.
The lawmakers called on AISI to assist industry initiatives to thwart PRC theft of AI technology and also devise strategies to respond to potential developments in the US-PRC AI rivalry. They concluded the letter by stating, "We look forward to engaging with the AI Safety Institute to better understand its ongoing work and resourcing needs on these topics," as noted in the SCCCP release.
Disclaimer: This post has been auto-published from an agency feed without any modifications to the text and has not been reviewed by an editor