The geopolitical battle over artificial intelligence has reached the doorstep of American enterprise. Two prominent U.S. House committees have formally requested information from food delivery giant DoorDash regarding its use of Kimi K2.6, a Chinese AI model developed by Beijing-based Moonshot AI. The investigation, reported by CNBC, marks a significant widening of the ongoing congressional probe into the security implications of U.S. companies relying on Chinese-developed artificial intelligence.
How DoorDash Triggered the Investigation
The inquiry was triggered by a public post on X by DoorDash co-founder Andy Fang, in which he detailed how the company was delegating certain lower-level AI workloads to Kimi K2.6. DoorDash’s AI research lab separately noted on X that Kimi K2.6 and Anthropic’s Fable 5 had vastly outperformed other models tested, including “Sonnet 4.6 and Opus 4.8 harness at a cheaper cost.” This public acknowledgment of relying on a Chinese model for core business functions immediately drew the attention of lawmakers.
The joint investigation is being conducted by the House Committee on Homeland Security and the House Select Committee on the Chinese Communist Party. In a letter obtained by CNBC, the committee chairmen asked DoorDash to share “information and documents” relating to its evaluation and deployment of AI systems from China. DoorDash responded diplomatically, stating it “proudly supports American AI leadership” and would engage with the committees on its responsible AI use.
A Systematic Congressional Effort
The DoorDash subpoena is not an isolated incident. It follows similar inquiries directed at Cursor and Airbnb, establishing a clear pattern of systematic congressional mapping of Chinese AI penetration into the American corporate ecosystem. The committee aide told CNBC that the investigation aims to determine whether the U.S. has a sufficient open-weight AI strategy so American companies are not “forced to choose between expensive or restricted U.S. models and cheap, capable PRC-developed alternatives.”
The core concern for policymakers is multifaceted. While open-weight models like Kimi K2.6 can be downloaded and run locally, mitigating some direct data privacy risks, lawmakers remain apprehensive about the underlying architecture, training data, and potential vulnerabilities in systems originating from China. House Homeland Security chairman Andrew Garbarino previously warned CNBC that a Chinese open-weight model matching U.S. models in “vulnerability discovery and cybersecurity tasks is highly alarming.”
The Impossible Tradeoff for American Business
The investigation places American companies in a genuinely difficult position. The rapid advancement of Chinese AI models, particularly in the open-weight category, has created a compelling value proposition. Models like Kimi K2.6 offer performance comparable to leading Western models at a fraction of the cost — a critical consideration for companies in margin-sensitive industries like food delivery. The political climate, however, is increasingly hostile to such pragmatism.
This tension is further complicated by the fact that the most capable open-weight models globally are currently Chinese-made. The recent incident in which a Chinese AI model was used to stop a rogue OpenAI agent’s attack on Hugging Face underscored the irony: American cyber defenders may sometimes have no choice but to rely on Chinese AI tools. The outcome of the DoorDash investigation will likely shape the future regulatory landscape for AI in the United States, drawing a clearer line between commercial strategy and national security risk.
What makes this investigation particularly consequential is its potential to set a precedent for how the U.S. government treats open-weight AI models from China. Unlike cloud-based AI services, where data flows to foreign servers, open-weight models run entirely on domestic infrastructure. The legal and regulatory framework for restricting their use is far less clear-cut than for, say, TikTok or Huawei equipment.
If Congress moves to restrict or ban the use of Chinese open-weight models by American companies, it would represent a significant escalation in the AI component of the broader tech war, and could accelerate the bifurcation of the global AI ecosystem into distinct Western and Chinese spheres. For companies like DoorDash, the immediate cost of compliance with any such restriction would be a return to more expensive domestic alternatives at a time when margins are already under pressure.
The investigation also raises a deeper question about the nature of technological competition. The United States has historically benefited from an open global technology ecosystem in which the best tools, regardless of origin, were freely available to American companies and researchers. Restricting access to Chinese AI models would mark a departure from that tradition, one with unpredictable consequences for American innovation.
The irony is that the congressional pressure to restrict Chinese AI may ultimately serve Chinese interests by forcing American companies to use less capable, more expensive domestic alternatives, ceding the efficiency gains that Chinese AI currently offers. This is the paradox at the heart of the DoorDash investigation, and it has no easy resolution. For context on the broader pattern of congressional scrutiny, see EastFrontier’s earlier coverage of Washington’s probe into US firms using Chinese AI models and the Moonshot Kimi K3 export controls allegation.