Two powerful US House committees have launched a formal investigation into DoorDash's deployment of a Chinese-developed artificial intelligence model, warning that the growing adoption of Chinese AI by American companies poses significant national security, cybersecurity, and economic security risks. The probe, announced on Friday, 1 August, marks a direct escalation of congressional scrutiny over China-linked technology embedded in mainstream US consumer platforms.

What Triggered the Investigation

The House Select Committee on China and the House Committee on Homeland Security jointly wrote to DoorDash co-founder and Chief Executive Officer Tony Xu, seeking detailed information about the company's evaluation and deployment of AI models developed in China. The move expands a broader congressional investigation that began in April and has already examined other American companies' use of Chinese open-weight AI models.

The committees cited public statements by DoorDash co-founder Andy Fang indicating the company had incorporated open-weight models into an internal AI-assisted code review system. According to the committees, DoorDash routed more demanding tasks to a US-developed frontier model while assigning lower-level work to Kimi K2.6, an open-weight model developed by China's Moonshot AI. The company reportedly stated that the new configuration outperformed an earlier setup while also lowering costs.

What the Committees Said

House Select Committee on China Chairman John Moolenaar and House Homeland Security Committee Chairman Andrew R. Garbarino acknowledged the commercial appeal of Chinese open-weight models in their letter. 'The Committees recognise that US companies, from large technology firms to startups, may evaluate and deploy PRC-developed open-weight models because they can provide competitive capabilities, lower costs, greater customisation, and alternatives to reliance on a small number of proprietary model providers,' they wrote.

However, the chairmen were unequivocal that commercial logic does not override security obligations. 'Those practical considerations do not eliminate the need for risk-based safeguards or diminish the national security concerns associated with growing dependence on models developed by entities subject to PRC jurisdiction,' the letter stated. The committees said the investigation aims to 'establish the factual record necessary to inform' a federal policy approach that would reduce US companies' reliance on Chinese-developed models while strengthening American open-weight alternatives.

The Moonshot AI Angle

Lawmakers also referenced allegations by the White House Office of Science and Technology Policy that Moonshot AI operated a covert platform to conduct large-scale distillation against American AI models, and that it trained models using advanced computing systems it was not authorised to obtain. These claims, if substantiated, would significantly deepen the security case against deploying Moonshot AI's products in US enterprise environments.

Why Open-Weight Models Are Under the Spotlight

Unlike proprietary AI systems, open-weight models make their underlying parameters publicly available, allowing developers to download, modify, and deploy them independently. The committees noted that while this flexibility reduces costs and increases customisation, it also raises concerns about software integrity, supply chain security, and the provenance of the underlying technology. Chinese open-weight models have become increasingly attractive to businesses seeking lower operating costs and the ability to run AI on their own infrastructure rather than through proprietary cloud services.

Broader Context and What Comes Next

This investigation fits into a wider pattern of Washington intensifying scrutiny of Chinese AI developers amid escalating strategic competition with Beijing over advanced technologies. The committees have signalled that their findings could form the basis for new federal safeguards specifically targeting high-risk applications. Industry observers note that DoorDash is unlikely to be the last company called to account — the probe's April origins suggest a systematic, sector-wide sweep is already underway.