Trump Walks Back Stricter AI Rules After Tech Giants Push Back
The United States has taken a notably lighter touch on artificial intelligence oversight after the tech industry mounted significant opposition to an earlier, more restrictive proposal.
President Donald Trump signed a revised executive order on AI governance on June 2, 2026 — one that strips out mandatory prerelease government reviews of advanced AI models, replacing them with a voluntary framework. The move marks a significant retreat from what had been floated as one of the most consequential federal interventions into the AI industry to date.
What Changed — and Why
The original draft order would have required AI developers to submit powerful new models for government evaluation before releasing them to the public. The idea was to give federal agencies a window to assess risks — from national security implications to potential misuse — before cutting-edge systems reached consumers and businesses.
But major players in the AI industry, including companies building large language models and AI infrastructure, objected strenuously. Their argument: mandatory prerelease reviews would slow innovation, create bureaucratic bottlenecks, and put American AI companies at a competitive disadvantage against rivals in China and elsewhere who face no such constraints.
The Trump administration, which has broadly favoured deregulation and positioned itself as pro-tech, ultimately sided with industry. The signed version of the order keeps the door open for government agencies to review AI systems — but participation is now entirely at the discretion of the companies involved.
Voluntary vs. Mandatory: A Meaningful Difference
Critics of the revised order argue that voluntary frameworks have a poor track record in tech policy. Without teeth, guidelines tend to become box-checking exercises rather than genuine safeguards.
Proponents counter that the AI landscape is moving too fast for government review processes to keep pace, and that heavy-handed regulation risks entrenching incumbents while freezing out startups that can't afford lengthy compliance procedures.
The debate echoes earlier battles over internet and social media regulation, where calls for oversight consistently ran into arguments about innovation and free expression — often with industry winning out.
The Broader Context
The executive order comes at a pivotal moment. AI capabilities have advanced dramatically over the past two years, with models now capable of writing code, generating synthetic media, conducting scientific research, and automating complex tasks at scale.
Governments around the world are grappling with how — and whether — to regulate the technology. The European Union has moved forward with its AI Act, one of the most comprehensive regulatory frameworks yet attempted. China has introduced its own set of AI governance rules. The United States, under the Trump administration, appears to be charting a deliberately hands-off course.
For AI developers, the voluntary framework is largely a win. For those concerned about the rapid deployment of powerful systems without independent safety checks, it represents a missed opportunity.
What Comes Next
The order leaves open the question of what happens when voluntary cooperation isn't enough. If a major AI incident — a serious misuse, a security breach, or an unforeseen harmful output at scale — occurs in the absence of mandatory oversight, the political calculus could shift quickly.
For now, however, the message from Washington is clear: the U.S. government will not be standing in the way of AI development.
Source: TechCrunch
