The White House is exploring plans to create an independent regulator dedicated to policing artificial intelligence safety, a move that could reshape how the federal government oversees frontier AI models from companies like OpenAI and Anthropic. The proposal, first reported by Bloomberg News, follows mounting complaints from Silicon Valley about the government's recent and sometimes unpredictable restrictions on new AI systems.
A FINRA for AI
According to sources familiar with the matter who spoke with Bloomberg, the plan is currently being reviewed by White House Chief of Staff Susie Wiles. It would establish an independent regulatory body focused specifically on artificial intelligence, reporting to the Securities and Exchange Commission, in a structure modeled on the Financial Industry Regulatory Authority, or FINRA.
FINRA is a self-regulatory organization that oversees brokerage firms and exchange markets in the United States. Translating that model to AI would mean creating an entity with the authority to set standards, conduct examinations, and enforce compliance among companies developing advanced models. The goal, according to Bloomberg's sources, is to bring consistency and predictability to an oversight process that has frustrated both tech companies and Wall Street firms.
The proposal comes at a moment of extraordinary tension between the federal government and the AI industry. Last month, the government issued export controls against Anthropic that forced the company to temporarily disable its Fable 5 and Mythos 5 models. Separately, OpenAI was compelled to make major changes to its newest Sol model before release at the government's request. Both companies protested the decisions, arguing they were excessive given the safety issues at stake.
Why the Push for Structured Oversight
The White House's interest in a dedicated AI regulator reflects a growing recognition that the current ad-hoc approach to oversight is unsustainable. Under the existing framework, the government has used a combination of executive orders, export controls, and voluntary agreements to influence how and when AI companies release their most powerful models. But this patchwork has drawn criticism from multiple directions.
Tech companies argue that inconsistent oversight makes it difficult to plan product roadmaps, as rules can change suddenly and without clear procedural standards. Wall Street firms, meanwhile, have pressed the government to take stronger action on AI-driven cybersecurity risks, warning that unchecked model capabilities could enable sophisticated attacks on financial infrastructure.
Bloomberg's sources said the proposal is designed to address both concerns, providing more certainty for AI developers while establishing a formal mechanism for evaluating and mitigating risks. By creating a FINRA-like entity, the administration hopes to shift from reactive interventions to a proactive regulatory regime with defined processes and accountability.
Echoes of Hassabis's Call for Standards
The White House deliberations come just days after Google DeepMind CEO Demis Hassabis publicly called for a United States-led standards body to independently test advanced AI models for national security threats before their release. In remarks reported by the Financial Times, Hassabis warned that artificial general intelligence capable of matching the human brain across a broad range of tasks is probably only a few short years away, leaving society a precious window to establish oversight.
Hassabis specifically proposed modeling the organization on FINRA, the same framework now under consideration inside the White House. He suggested that a U.S.-initiated system could become the basis for international standards as governments confront risks from increasingly capable AI systems.
The convergence between Hassabis's proposal and the White House's internal deliberations suggests growing consensus around the FINRA model, even as the details of scope, authority, and enforcement remain to be negotiated.
Existing Prerelease Oversight
The government has already taken steps toward prerelease oversight. An executive order issued in June sought voluntary access to covered frontier models for up to 30 days prior to release, allowing federal officials to gauge advanced cyber capabilities. Google DeepMind, Microsoft, and xAI all agreed to submit models for federal national security testing under the framework.
However, these measures remain voluntary and limited in scope. A dedicated regulator would have far broader authority to compel transparency, mandate safety testing, and enforce compliance, fundamentally altering the relationship between AI companies and the federal government.
What Comes Next
The proposal is still in its early stages, and there is no guarantee it will advance in its current form. Creating a new regulatory entity requires congressional action, and the political dynamics surrounding AI regulation remain fluid. Lawmakers have held numerous hearings on AI safety but have yet to pass comprehensive legislation.
Still, the fact that the White House is actively weighing a FINRA-style regulator signals a shift in ambition. Rather than relying on executive orders and case-by-case interventions, the administration appears to be exploring a more permanent institutional framework for AI oversight, one that could outlast any single administration and provide the consistency that industry leaders have been demanding.
For AI companies, the prospect of a dedicated regulator is a double-edged sword. It could bring the predictability they have asked for, but it could also introduce new compliance burdens, reporting requirements, and delays in bringing products to market. How those tradeoffs are balanced will shape the competitive landscape of the AI industry for years to come.
Stay Ahead of AI
For ongoing coverage of AI regulation, government policy, and the oversight frameworks shaping the industry, follow AI Buzz Wire for reporting on the decisions that matter.
Read more AI policy news →
