Trump Administration Weighs FINRA-Style Self-Regulatory Body for AI

If adopted, a private self-regulatory organization modeled on financial industry precedent would place primary AI oversight outside direct congressional or executive agency control, raising...

The Trump administration is considering the creation of a self-regulatory organization (SRO) for frontier artificial intelligence modeled on the Financial Industry Regulatory Authority, according to reporting by the Washington Examiner, published September 3, 2026. No formal legislative proposal or executive order has been made public as of this date.

FINRA, the Financial Industry Regulatory Authority, is a private, not-for-profit corporation authorized by Congress under the Securities Exchange Act of 1934 and reorganized under its current structure in 2007. It oversees approximately 3,400 broker-dealer firms and around 620,000 registered securities representatives, according to FINRA's own published 2024 annual report. FINRA is not a federal agency; it derives its authority from a delegation by the Securities and Exchange Commission.

The SRO model, if applied to AI, would mean that a private body, rather than a federal agency such as the Federal Trade Commission or the National Institute of Standards and Technology, would set binding standards and conduct enforcement for companies developing or deploying advanced AI systems. Which companies would be subject to membership, what penalty powers the body would hold, and who would appoint its leadership are all unknown. The document that would answer those questions is a formal legislative proposal or executive order, neither of which has been released.

Proponents of the FINRA model for AI argue that technical complexity in frontier AI parallels the complexity of securities markets, where industry practitioners possess expertise that government regulators may lack. The Washington Examiner commentary notes that FINRA maintains a budget funded by industry fees and exercises quasi-governmental authority over member firms, including the power to bar individuals from the industry.

Critics of the SRO model, including the Washington Examiner's own prior editorial stance from September 2025, have characterized existing SROs as opaque private bodies exercising powers that belong in government. That earlier editorial called FINRA and similar organizations an administrative state inside the administrative state. The phrase is descriptive of a structural concern: SRO rulemaking and disciplinary proceedings are not subject to the same notice-and-comment requirements under the Administrative Procedure Act that apply to federal agencies.

Congress has not passed any comprehensive federal AI legislation as of September 2026. The AI Act introduced in the 118th Congress did not advance to a floor vote. The National AI Initiative Act of 2020 authorized coordination across federal agencies but did not create a regulatory body with enforcement power. No equivalent legislation has been enacted in the 119th Congress through the date of publication.

The FTC retains existing authority to pursue unfair or deceptive practices under Section 5 of the FTC Act, which legal scholars have argued could apply to certain AI deployments. The agency has brought enforcement actions involving algorithmic systems, though none have established binding sector-wide AI standards. A congressional mandate or executive order would be required to establish any new SRO with delegated authority over AI.

The precedent for industry-created regulatory bodies in the United States extends beyond finance. The Nuclear Energy Institute and the North American Electric Reliability Corporation (NERC) both serve SRO-adjacent functions in their sectors, with NERC holding FERC-approved mandatory reliability standards authority under the Energy Policy Act of 2005. In each case, the scope of private authority was defined by an explicit statutory delegation.

For an AI SRO to hold enforceable authority, Congress would need to pass enabling legislation, or a federal agency would need to formally delegate rulemaking and enforcement functions. The legal mechanism, the agency involved, and the scope of covered technologies remain unknown. The FTC Act, the Commerce Department's existing AI Safety Institute authorization, and potential new legislation are among the vehicles that could be used.

What remains publicly unknown includes the specific agencies involved in the administration's deliberations, the names of any private entities participating in preliminary discussions, the proposed funding mechanism, and whether any draft legislation or executive order text exists. A Freedom of Information Act request directed to the Office of Science and Technology Policy or the National Economic Council, combined with any future LDA lobbying disclosures filed by AI companies engaged in these discussions, would be the most direct public records to watch.