Regulatory  ·  2026-08-05

House Select Committee on China and Homeland Security Committee expand probe into corporate use of Chinese AI models (DoorDash)

RegulatoryMedium impactUnited States
House Select Committee on China chairman John Moolenaar and House Homeland Security Committee chairman Andrew Garbarino sent a letter (reported through Aug 3-4, 2026 coverage) to DoorDash CEO Tony Xu demanding disclosure of which Chinese AI models (e.g., Moonshot AI's Kimi K2.6) the company uses, what data is exposed to them, and what security testing was performed, with a briefing deadline of August 14 and document deadline August 21. This extends an earlier joint investigation (announced April 29, 2026) into PRC-developed models including DeepSeek, Alibaba, Moonshot AI and MiniMax, having already contacted Airbnb and Anysphere (Cursor).
Signals escalating congressional scrutiny of enterprise use of open-weight Chinese AI models over 'hidden vulnerability' and national-security concerns, potentially foreshadowing legislative restrictions or disclosure mandates on companies that integrate Chinese-origin models into US commercial products.
US companies using Chinese-origin AI models (open-weight or API-based) should document model provenance, data-handling, and security testing in case of similar congressional inquiry; monitor for potential legislative follow-through.
South China Morning PostLexology
See this in the live feed Explore related AI security and governance findings — updated every morning.
Open the feed →