U.S. Investigates Security Risks from Chinese AI Models in Airbnb and Anysphere

On April 29, 2026, House Committee leaders announced a joint investigation into national security and cybersecurity risks associated with the use of Chinese-developed artificial intelligence models by U.S. companies Airbnb and Anysphere. The investigation is prompted by concerns that these AI models, developed by companies such as DeepSeek and Moonshot AI, could introduce vulnerabilities and expose sensitive data to the Chinese government. The Chairmen highlighted that these models are trained under China's censorship regime, potentially compromising critical infrastructure in the U.S. The inquiry follows a White House memo warning of industrial-scale campaigns by foreign entities to distill U.S. AI systems. Initial letters were sent to both companies, focusing on the risks posed by the Cursor’s Composer 2 model linked to Moonshot AI. The investigation aims to assess the implications of adopting these AI technologies on American innovation and security.

Key Points: • House Committee leaders announced an investigation into Chinese AI models used by Airbnb and Anysphere. • Concerns include potential data exposure to the Chinese government and vulnerabilities in critical infrastructure. • The investigation follows a White House memo on foreign campaigns targeting U.S. AI systems.