Skip to content

Congress Expands Investigation into DoorDash's Adoption of PRC AI

By News Analysis India · · 2 min read

On August 1, senior members of the House Select Committee on China and the Homeland Security Committee announced an expanded probe into DoorDash’s use of artificial‑intelligence models that originate from the People’s Republic of China. The committees dispatched a formal request to DoorDash co‑founder Tony Zhu, seeking comprehensive data on the evaluation process, licensing, and practical deployment of Chinese‑developed AI within the company’s services.

The inquiry extends the scope of a wider congressional audit that began in April, which had already scrutinized American firms’ adoption of Chinese open‑weight AI frameworks. Committee chairs John Moolenaar and Andrew Garbarino noted that cost efficiency, customizability, and competitive performance drive U.S. firms toward these models, but warned that such benefits cannot eclipse the security implications.

In their letter, the chairs emphasized that reliance on PRC‑created AI does not eliminate the necessity for risk‑based safeguards and that growing dependence raises alarms for national‑security, cyber‑defense, and economic resilience. They urged the development of a robust federal response that bolsters U.S. open‑weight alternatives, enforces stringent security standards, and preserves competition.

Key investigation points include the integration of Chinese AI into consumer‑facing platforms, enterprise applications, and systems that process sensitive commercial or personal information. The committees aim to assess whether such integration could compromise software integrity, supply‑chain security, or expose proprietary technology provenance.

DoorDash’s public reply clarified that its internal AI‑assisted code‑review workflow combines a domestically engineered “Frontier” model for complex tasks with a lower‑tier model, “kimi‑2.6,” supplied by China’s Moonshot AI, for simpler operations. According to the company, this hybrid approach has delivered better performance and lower costs compared with earlier configurations.

Congressional staff also referenced remarks from the White House Office of Science and Technology Policy, which accused Moonshot AI of running a secret platform that performed large‑scale distillation of U.S. AI models using advanced computing resources without authorization.

Chinese open‑weight AI platforms are rapidly attracting interest for their reduced operating expenses, flexibility, and the ability to run on an organization’s own infrastructure rather than relying on proprietary cloud services. As strategic competition with Beijing intensifies, U.S. policymakers are accelerating efforts to examine and potentially curb the influence of Chinese AI technology on American businesses.