Leading Democrat asks AI firms for data on any Chinese access to sensitive code | Reuters

A Democratic lawmaker has asked five leading US artificial intelligence companies for data on any Chinese access to their sensitive code, according to letters seen by Reuters. The letters were sent to the CEOs of OpenAI, Anthropic, Google, Meta and SpaceXAI. Reuters reported on the request on October 1, 2026, and describes the sender only as a leading Democrat.

In the letters, the lawmaker warns that Beijing could steal AI model weights developed by OpenAI, Anthropic and other top US firms. The lawmaker wrote that the theft of an advanced model weight by a hostile non-state actor 'could endanger all of humanity.' The article names no specific incident in which model weights are alleged to have been taken.

The letters are a request for information rather than a regulatory or legal action. The article centres on model weights and the risk that a foreign adversary could obtain them. It cites both Beijing and 'a hostile non-state actor' as potential sources of that risk. The article states that the loss of an advanced model weight could affect the US position relative to China in AI development, and frames the matter as part of the wider US–China competition over advanced artificial intelligence.