US lawmakers urge keeping Chinese open-weight AI models out of federal systems and critical infrastructure

Two House committees found Chinese open-weight AI models are widely used by US companies and made 12 recommendations, including keeping them out of government work.

21 Nov 20261 min readModelsPolicyChinaUnited States
12: recommendations from two House committees, including keeping Chinese models out of federal systems

Chinese-made AI models that anyone can download and adapt should be kept out of US federal systems, government contract work and high-risk parts of critical infrastructure, two US House committees said on 20 November. The House Homeland Security Committee, chaired by Andrew Garbarino, and the House Select Committee on China, chaired by John Moolenaar, published a joint report after an investigation that began in April.

The committees wrote to three US companies. They say Airbnb “uses Alibaba models in customer support”, Anysphere built the flagship model in its Cursor coding tool “on a Moonshot AI base model”, and DoorDash “reported that multiple PRC model families had been authorized and deployed.” The report finds that “cost, customization, and near-frontier performance are driving adoption”, and argues that much of China’s apparent progress “rests on capabilities extracted from American models.”

It makes 12 recommendations, including a national programme to build American open-weight models, a ban on Chinese-developed models in federal systems, and action against “illicit distillers” who copy US models.

Why it matters

open-weight models are cheap and popular with businesses, so a federal ban could change which AI tools contractors and infrastructure operators are allowed to use. For now this is a committee report, not a law, and no bill was announced alongside it.


Related