US lawmakers have launched a probe into Chinese artificial intelligence firms, citing potential national security risks and allegations of exploiting American AI models. The investigation, led by the House Select Committee on the Strategic Competition Between the United States and the Chinese Communist Party, targets companies including Baidu, Tencent, and SenseTime. The committee has requested documents and information regarding the firms' access to US technologies and their compliance with export controls.
Allegations of Model Theft
The probe focuses on claims that Chinese entities have reverse-engineered or directly copied proprietary AI models developed by US companies. Lawmakers are particularly concerned about the use of American open-source models that may have been adapted without proper attribution or licensing. The investigation also examines whether these firms have violated sanctions by providing AI tools to entities in Iran, North Korea, or other sanctioned nations.
National Security Implications
The committee argues that the transfer of advanced AI capabilities to China could undermine US technological leadership and pose risks to critical infrastructure. AI systems developed with stolen or misappropriated technology could be used for surveillance, disinformation, or military applications. Lawmakers are also scrutinizing the role of US venture capital firms that have invested in Chinese AI startups, potentially facilitating technology transfer.
In response, Chinese officials have dismissed the allegations as baseless and accused the US of engaging in protectionism. The Chinese Ministry of Foreign Affairs stated that China is committed to ethical AI development and opposes any form of technology theft. Meanwhile, the investigated companies have denied wrongdoing, emphasizing their compliance with international laws and regulations.
The probe marks the latest escalation in US-China tensions over technology, following previous actions against Huawei and TikTok. The committee plans to hold hearings later this year to discuss the findings and consider potential legislative measures to safeguard US AI assets.
