HR-10152-119
Ordered to be Reported by the Yeas and Nays: 26 - 21.
Sponsored by Gabe Evans (R-CO)
What it does
This bill would direct the Secretary of Commerce to take actions supporting the adoption of "qualified open" AI models developed by U.S. persons, including coordinating with industry, identifying barriers, and entering agreements with private companies, states, and foreign partners. It would also require Commerce to annually assess and publicly report on risks posed by "foreign adversary" AI models from countries like China, while explicitly barring the Secretary from banning or restricting any open AI model.
Who benefits
U.S.-based AI companies and startups developing open-weight models, who could gain federal coordination support, agreements, and promotional assistance from Commerce. Domestic AI hardware and cloud providers indirectly, and U.S. entities seeking alternatives to Chinese-developed AI models. State governments and federal agencies that adopt qualified open models may gain Commerce Department technical partnership.
Who is hurt
Developers and distributors of foreign adversary AI models (e.g., Chinese-linked open-source models) could see reduced U.S. institutional adoption due to the risk reporting, potentially affecting their U.S. market share. Companies building closed/proprietary AI models are not directly supported by this bill and could see relatively less federal promotional attention. Taxpayers bear the modest administrative cost of the new coordination office and reporting requirements.
Supporters argue
Supporters argue that open-source AI models developed in the U.S. strengthen national competitiveness and transparency compared to closed foreign systems, and that a designated federal point of contact would reduce bureaucratic barriers to adoption. They contend that publicly assessing risks from foreign adversary models, particularly from covered nations like China, gives businesses and government agencies better information to make security-conscious choices without banning any technology outright.
Opponents argue
Opponents argue that directing Commerce to affirmatively promote a specific category of private companies' products risks favoritism and blurs the line between neutral regulation and industrial policy, potentially disadvantaging competing AI development approaches. They contend that the annual "foreign adversary model" risk reports could be used to informally pressure institutions away from certain technologies without the due process protections that would accompany a formal restriction, even though the bill explicitly bars outright bans.
Constitutional context
Congress's regulation of AI models in interstate and foreign commerce rests on the Commerce Clause (Art. I, §8, cl. 3), and the bill's rule of construction avoiding any ban on open models sidesteps First Amendment concerns about compelled or restricted speech that arose in Moody v. NetChoice (2024) regarding platform and content regulation.
Checks and balances
The executive branch (Commerce Department) gains new coordination and reporting authority, but the bill itself contains no enforcement power, cannot restrict any model, and must report findings to congressional committees, preserving legislative oversight.
Historical precedent
The bill's structure resembles other sector-specific federal coordination and risk-assessment mandates, such as CHIPS Act provisions directing Commerce to support domestic semiconductor development, though no directly analogous open-source AI promotion law has been enacted.