US lawmakers urge AI Safety Institute to counter China's growing AI Threat, citing DeepSeek risks

Washington DC [US], June 4 (ANI): Chairman John Moolenaar (R-MI) and Ranking Member Raja Krishnamoorthi (D-IL) of the House Select Committee on the Strategic Competition Between the United States and the Chinese Communist Party (CCP) are calling on the Department of Commerce to enhance the responsibilities of the U.S. AI Safety Institute (AISI) to tackle the growing national security challenges posed by China's advancements in artificial intelligence, as stated in the release from the Select Committee on the Chinese Communist Party (SCCCP).
In a correspondence to Commerce Secretary Howard Lutnick, the legislators stressed that "there is a strong national security need for better understanding, predicting, and preparing for the PRC's AI progress." They pointed out the approaching January 2025 introduction of DeepSeek's large language model, R1, as a significant alert. The Committee's probe into DeepSeek uncovered "multiple national security risks, including the funneling of Americans' private data to the PRC, manipulation of the model's outputs according to PRC law, and the potential theft of U.S. AI technology through model distillation," as noted in the release.
As AI technologies advance further, the lawmakers cautioned, "the need to anticipate and prepare for PRC AI developments and to prevent strategic surprises will only become more pressing," according to Moolenaar and Krishnamoorthi, as cited by the SCCCP's release.
As per the SCCCP statement, the Committee has urged AISI to broaden its goals to address PRC AI growth, emphasising that "a comprehensive government strategy will be necessary to ensure sustained U.S. leadership in cutting-edge AI innovation." They also pointed out AISI's "distinct technical expertise, robust collaborations with the industry, and experience in testing and evaluation" as important resources in the U.S. initiative to remain ahead of the CCP.
Moolenaar and Krishnamoorthi specified certain domains where AISI could assist in safeguarding U.S. national security, including assessing the national security strengths and weaknesses of leading PRC AI models, comparing and evaluating PRC models against U.S. models to understand the competitive scenario, aiding industry efforts to thwart PRC theft of AI technology, and formulating responses to potential shifts in the U.S.-PRC AI rivalry. The lawmakers concluded the letter by stating: "We look forward to collaborating with the AI Safety Institute to better comprehend its ongoing projects and resource requirements regarding these issues," as mentioned in the SCCCP release. (ANI)