The White House has a draft. It has been circulated. It has not moved. That is the entire story, and it is a loud one. The proposed executive order to establish a self-regulatory body for AI—an industry-led SRO with federal backing—has stalled internally, according to multiple sources. No formal announcement. No timeline. No explanation. Just silence from the West Wing.
Code does not lie, but it often omits the context. In this case, the code is policy, and the context is a power struggle. The stagnation is not a bureaucratic hiccup. It is a signal. The Trump administration has not unified on AI governance, and the window for a coherent federal framework is closing. Every week of inaction cements a future where fifty states write fifty different rules, and Brussels writes the global ones.
The Context: A Philosophical Reversal
To understand why this stall matters, you must first understand what the order was supposed to do. The Biden administration's October 2023 executive order was federal-first: multiple agencies, mandatory reporting, safety assessments. It was heavy, slow, and prescriptive. The Trump administration's approach was the inverse. Industry self-regulation. Voluntary compliance. A single SRO entity, chartered by the federal government but operated by the companies it oversees.
This is not a novel concept. The financial industry has FINRA, a self-regulatory organization that writes and enforces rules for broker-dealers. It works, imperfectly, because Congress explicitly authorized it. The AI version has no such legislative foundation. An executive order cannot grant private entities regulatory authority over their own competitors. That requires an act of Congress. This is the first crack in the foundation, and it is structural.
The philosophical shift is stark. Biden's order prioritized safety over innovation. The proposed Trump order prioritizes innovation over safety. On paper, that sounds like a pro-business stance. In practice, it creates a legal and political quagmire that the White House legal counsel has likely flagged. You cannot delegate regulatory power you do not constitutionally possess.
The Core: Why the Order Is Stuck
The stall is not a single point of failure. It is a distributed denial of service attack from multiple fronts.
Front One: Internal White House Division. The draft has been circulated but has not progressed. In policy terms, that means the National Security Council and the Commerce Department are at odds. The NSC wants stricter export controls and foreign investment review. Commerce wants a lighter touch to keep innovation flowing. The legal counsel is likely warning that an SRO without congressional authorization is a constitutional overreach. The political team is calculating the electoral cost of pushing a controversial regulatory reform in an election year. Three different priorities, one stalled document.
Front Two: The Tech Industry's Ambivalence. Publicly, tech giants support self-regulation. It is lighter than federal mandates. But privately, the math is more complex. An SRO dominated by OpenAI, Google, and Meta looks like a legalized cartel. That invites antitrust scrutiny. Smaller companies face higher compliance costs because the standards will be written by the incumbents. And there is liability risk: if you participate in writing the rules, you are more exposed when those rules fail. The industry wants deregulation, but it does not want to be the ones holding the pen when the next AI incident occurs.
Front Three: Congressional and State Pushback. Congress has its own AI bills. It does not want to cede authority to an executive-branch-created SRO. The states are already moving. California passed SB 53, requiring safety testing and transparency reports for large AI models. Colorado passed SB 205, the first comprehensive AI consumer protection law. New York City has Local Law 144 for AI hiring audits. At least forty states have introduced AI-related legislation. The federal vacuum is not empty. It is filling with state-level patchwork.
Here is the hidden detail most analysts miss: the draft order reportedly includes a preemption clause to limit states from passing conflicting rules. That is a direct assault on federalism. It is also a guaranteed legal challenge. The Supreme Court has been skeptical of broad federal preemption in recent terms. The White House knows this. The clause is likely a poison pill that the legal team cannot swallow.
The Contrarian Angle: Strategic Stagnation
What if the stall is not a failure but a strategy? In an election year, controversial regulatory architecture is a liability. The White House may be deliberately parking this order until after November. If the administration wins, it can push the SRO framework with a fresh mandate. If it loses, the problem belongs to someone else. This is not incompetence. It is risk management.
But there is a darker interpretation. "Self-regulation" in this context may be a euphemism for deregulation. The order, if enacted, would likely dismantle the Biden-era safety assessment and reporting obligations. The stall may be a quiet victory for the industry: no new federal rules, no mandatory reporting, no safety assessments. The status quo is the desired outcome. The order is a decoy.
Based on my experience auditing protocol governance structures, I see a parallel. In DeFi, a governance token that never gets used for a vote is not a failure. It is a mechanism for maintaining the illusion of decentralization while preserving centralized control. The stalled executive order functions the same way. It signals action without delivering change. The market reads this as stability. The industry reads it as a green light.
The Takeaway: The Vacuum Fills Itself
The most likely outcome is not a revived federal framework. It is a permanent state of fragmentation. California's SB 53 will become the de facto national standard because any company selling AI products in the largest state economy must comply. The EU AI Act, which took effect in August 2024, will become the global standard because it is the only comprehensive framework with enforcement teeth. The "Brussels Effect" is real. GDPR proved it. AI regulation will repeat the pattern.
The United States is not just losing the AI governance race. It is not competing. Every month of federal inaction is a transfer of rule-making authority to Sacramento and Brussels. The question is not whether the executive order will pass. It is whether the United States will have any seat at the table when the global standards are written. The current answer is no.
Watch the Q4 signals. If the order resurfaces after the election, the stall was tactical. If it remains buried, the stall was substantive. Either way, the window for a unified federal framework is closing. The states are not waiting. The EU is not waiting. The market is not waiting. The only question is who writes the rules you will eventually have to follow.
