House Republicans have opened a formal investigation into the use of Chinese-developed artificial intelligence by American companies, targeting Airbnb and Anysphere, the parent company of the coding tool Cursor. The probes, announced Thursday, mark the first step in a broader congressional review of national security risks tied to Chinese AI models embedded in U.S. products.
In letters sent to both firms, the chairmen of the House Homeland Security Committee and the House Select Committee on China warned that Chinese AI systems could pose a direct threat to U.S. national security. The lawmakers, Rep. Andrew Garbarino (R-N.Y.) and Rep. John Moolenaar (R-Mich.), are demanding detailed disclosures about any Chinese models used, along with internal security assessments and customer data exposure risks.
The investigation zeroes in on Cursor's new model, Composer 2, after an independent developer found it was built on systems from Moonshot AI, a Beijing-based startup backed by Alibaba Group. According to the lawmakers' letter, Moonshot AI is one of three Chinese firms that have run "coordinated campaigns to extract advanced capabilities" from American AI systems. The other two are DeepSeek and MiniMax.
“When capabilities are stripped out through distillation and repackaged without equivalent safeguards, the resulting models may become available to hostile state actors, terrorist organizations, and criminal enterprises,” the lawmakers wrote. Distillation is a technique where outputs from a more powerful model are used to train a less capable one, allowing foreign actors to release products that appear comparable at a lower cost while lacking the original's full performance.
The Republican chairs also pressed Airbnb about its reported use of Alibaba’s Qwen large language model in customer service operations. They cited comments by CEO Brian Chesky, who last year said the company is “relying a lot” on Qwen, calling it “very good” and “fast and cheap.”
“Alarmingly, this conduct is part of a broader PRC campaign to accelerate its AI capabilities through the exploitation of American innovation, including through espionage, intellectual property theft, and other unlawful or deceptive means,” the lawmakers wrote.
The Hill reached out to Airbnb and Anysphere for comment. The push comes nearly a week after White House Office of Science and Technology Director Michael Kratsios warned in a memo that Chinese entities are using thousands of proxy accounts and jailbreaking tactics to access proprietary U.S. information.
The investigation underscores ongoing tension in the AI race between Washington and Beijing. The Trump administration has faced backlash from its own party for allowing Nvidia to sell its H200 chips to China, which are more powerful than the H20 chips designed to comply with U.S. export controls. After initially restricting sales, the administration reversed course and permitted the sales in exchange for a 15 percent cut of revenue.
Each company must disclose all Chinese AI models deployed in their operations, provide any independent security evaluations, and submit internal analyses and communications related to the models. The lawmakers also asked for details on any customer data that could be at risk.
The investigation adds to a growing list of bipartisan concerns over Chinese technology infiltration, including efforts to bar Chinese automakers from the U.S. market and scrutiny of partnerships like Senator Bernie Sanders’ collaboration with a Chinese AI official. Meanwhile, the administration continues to target Chinese refineries and shipping to cut Iran oil revenue.
