Trump delays AI security review order to avoid slowing U.S. innovation

President Donald Trump delayed signing an executive order on artificial intelligence this week after saying he was uneasy with the language in the document and worried it could slow U.S. progress in the fast-moving technology race. According to TechCrunch and Fast Company, the order would have required government security reviews before some AI models were released, but Trump decided not to move forward with it ahead of an expected White House ceremony.

Trump said he did not want to “get in the way” of American leadership in AI and suggested the proposed rules might have become a “blocker.” In remarks reported by Fast Company, he said he did not like what he was seeing in the draft and was concerned it could weaken the country’s advantage over competitors such as China. The president has repeatedly framed AI development as a strategic priority for the United States.

The delayed order appears to have been aimed at addressing national security risks tied to increasingly powerful AI systems, including possible misuse or harmful outputs. But the administration’s hesitation underscores a familiar tension in Washington: how to encourage rapid innovation while still imposing safeguards on a technology that can be used for cyberattacks, misinformation, surveillance, and other risks. Supporters of tighter oversight argue that some form of review is necessary before advanced models are widely deployed, while critics say heavy-handed rules could push development overseas.

The move also fits with Trump’s broader approach to AI policy, which has emphasized expanding the U.S. industry and reducing what he sees as unnecessary barriers. His administration has already taken steps in that direction, including orders focused on boosting AI development and limiting what it has described as burdensome regulation. The new delay suggests the White House is still trying to balance security concerns with its desire to keep the U.S. ahead in the global AI race.

For AI companies, the decision postpones what could have been a new layer of federal scrutiny over model releases. For lawmakers and regulators, it leaves unresolved a central question: whether the government should establish clear pre-deployment safety checks for advanced AI systems or continue relying on a lighter touch that leaves more room for innovation. What happens next will likely depend on whether the administration revises the order, abandons it, or returns with a narrower version.