A conservative coalition led by Steve Bannon demands a Trump executive order requiring government testing and approval of frontier AI before any commercial release.
Steve Bannon and 60 co-signatories delivered a letter to President Trump last Friday demanding that frontier artificial intelligence systems clear a federal approval process before reaching the public. The letter, organized by Humans First, a self-described conservative social movement, requests an executive order mandating testing, evaluation, vetting, and government sign-off for advanced AI models prior to any commercial release.
Bannon's position, as he told Axios, was unambiguous: mandatory government approval is a "must." Among the letter's signatories are Amy Kremer, Humans First chairwoman and longtime Trump supporter, and approximately three dozen church leaders.
The alarm is broad. According to the Fox Baltimore report, the letter flags potential AI threats across cybersecurity, critical infrastructure, financial systems, election integrity, and biosecurity, along with vaguer concerns about children, workers, and what the signatories call threats to the American "way of life." No specific system or company is named. The coalition's framing is explicitly nationalist: "America did not become the greatest nation in the world by allowing unelected elites to experiment on the public without safeguards or accountability."
The industry context
The demand lands at an awkward moment for AI developers. ChatGPT alone has reached 800 million monthly active users and $20 billion in annual recurring revenue, as Forbes reported earlier this year. A scale that large makes a mandatory approval process not just a policy question but a logistical one. Even a 30-day pre-release review regime, applied consistently, would reshape product roadmaps across the industry.
AI companies are simultaneously pushing harder to monetize. OpenAI launched its ChatGPT Go tier globally this year at $8 per month, part of a broader shift toward recurring revenue as investor patience for indefinite cash burn narrows, as the Digital Watch Observatory has tracked. A mandatory approval bottleneck would complicate the product-release velocity that subscription growth depends on.
What makes this letter different
Previous calls for pre-release AI audits have come mainly from academic researchers, civil-society groups, and former government officials, largely associated with the center or left. A letter bearing Bannon's name alongside roughly three dozen evangelical church leaders represents a different kind of pressure: culturally conservative, explicitly pro-America First, and aimed at an administration that has leaned toward deregulation in the AI sector.
The coalition does not oppose AI development. It explicitly says America should lead the world in AI innovation. The distinction it draws is between innovation and "unchecked power": the argument that no private corporation should have unilateral authority to release an untested system on the public. That framing may prove easier to sell inside the current White House than the precautionary language favored by academic AI safety researchers.
An executive order, if issued, would need to define what "frontier" means, which agency holds approval authority, and what technical standard systems must meet to clear review. None of those questions have settled answers. The National Institute of Standards and Technology has published a voluntary AI risk management framework, but voluntary guidance is far from a mandatory pre-release clearance regime with real enforcement teeth. An artificial intelligence review process with binding power would mark a significant policy shift from the current self-regulatory ecosystem.
Google's developer conference, I/O 2026, opens Tuesday and is expected to deliver a wave of AI product announcements, per CNET. If the Humans First letter gains traction in Washington, the timing gives major AI developers a concrete reason to engage publicly with the oversight debate rather than wait for it to arrive at their door.
Whether Trump acts is uncertain. The administration has pursued deregulation broadly, and AI companies have cultivated relationships with key White House figures. A formal review mandate would represent a sharp break from that posture. The more probable near-term effect is that the letter legitimizes the conversation and forces AI companies to prepare arguments against mandatory testing from a constituency they did not expect to face.
---
FAQ
What is Humans First?
Humans First describes itself as a conservative social movement. Chairwoman Amy Kremer is a political activist and longtime Trump supporter. The group sent the letter to the president on Friday, May 16.
What would a mandatory AI approval process look like?
The letter calls for an executive order requiring testing, evaluation, vetting, and government approval before any frontier AI system is released to the public. It does not specify which federal agency would administer reviews or what criteria systems must meet.
Which AI systems would be affected?
The letter targets "frontier AI models" without naming specific products or companies. Defining "frontier" in a regulatory context would be one of the central challenges of any executive order that follows.
Has the Trump administration signaled support for AI oversight?
The administration has generally favored deregulation of artificial intelligence. The Bannon letter applies pressure from within the president's own political coalition, which may be harder to dismiss than criticism from outside it.