- Proposal. Representative Pramila Jayapal (D-WA) said on September 17 that large AI companies should be required to obtain a national federal charter to operate, modeled on the banking system’s licensing regime.
- Stakes. The proposal directly challenges the AI industry’s preferred regulatory approach—seeking antitrust exemptions for intercompany collaboration rather than new structural requirements.
- Timing. The announcement came at a House hearing on AI surveillance by Big Tech, as companies including Anthropic and OpenAI press Congress for legislation on their own terms.
Representative Pramila Jayapal, Democrat of Washington and co-chair of the House Monopoly Busters Caucus, previewed a forthcoming bill that would compel large artificial intelligence companies to obtain a federal operating charter—a requirement analogous to the licensing regime that governs banks—as a condition of doing business. The announcement came on September 17 at a shadow hearing entitled “AI Crossroads: Stopping Big Tech’s Race to Monopolize AI through Surveillance,” which Jayapal convened alongside members of the Progressive Caucus, according to reporting from MLex.
What the charter framework would require
Under Jayapal’s outline, companies would need a charter to operate and the charter would be conditioned on structural reforms—a formulation deliberately designed to give regulators ongoing leverage over AI firms’ business models, not just a one-time approval. “The industry needs new regulation, not exceptions to existing law,” she said. The comment was a pointed rebuke of the AI sector’s active lobbying effort to secure an antitrust waiver that would allow competing labs—including OpenAI, Anthropic and Google—to collaborate on safety standards without triggering competition scrutiny.
The proposal echoes the approach California took earlier this year when it became the first US state to mandate third-party audits of high-risk AI systems—legislation that federal lawmakers have watched closely as a potential template. Jayapal’s framework would go further by establishing a structural entry requirement rather than a periodic compliance check.
The industry’s counter-position
Major AI labs have worked to shape federal legislation in a direction that would pre-empt stricter state rules and establish safety norms through voluntary intercompany dialogue rather than government-imposed structural requirements. Anthropic and OpenAI have both raised safety concerns about their own frontier models while simultaneously arguing that coordination between labs is necessary to manage catastrophic risk—a position critics say is designed to insulate incumbents from new entrants and from mandatory disclosure obligations.
The Department of Justice has said it remains open to dialogue on AI collaboration but acknowledged it is difficult to predict in advance whether such arrangements would violate antitrust law—a response that AI companies seeking certainty found insufficient. Jayapal’s charter bill, when formally introduced, is expected to be opposed by Republican members who have consistently favoured lighter-touch federal AI regulation to avoid burdening US labs competing with Chinese counterparts.
Legislative outlook
The charter bill is a preview rather than a filed piece of legislation, and its path in the current Congress is uncertain. The House went on recess after Speaker Mike Johnson cancelled further votes ahead of November’s midterm elections, meaning any floor action would have to wait for the next Congress. Jayapal’s hearing nonetheless served to anchor a clear Democratic position: AI regulation should impose structural obligations on the largest players, not grant them legal shelter from competition law.