Trump rejects new AI regulations as OpenAI and Anthropic call for oversight

The people building the most powerful AI systems on Earth are asking the government to regulate them. The government’s response, more or less: no thanks, we’ve got this.

President Donald Trump took to Truth Social on September 14 to reject growing calls from AI industry leaders for new regulatory guardrails on frontier AI development. His position was characteristically blunt: the only controls necessary come from a “strong and smart (High IQ!) President,” and safety concerns raised by industry figures amount to a “hoax” comparable to climate change warnings.

The labs want rules, the White House wants speed

The flashpoint was an essay from Anthropic CEO Dario Amodei advocating for a global slowdown in frontier AI development to address mounting risks. That essay didn’t land in a vacuum. It picked up endorsements from OpenAI CEO Sam Altman, Elon Musk, and Demis Hassabis of Google DeepMind, creating something close to an industry consensus that the current pace of advancement needs structured oversight.

Trump wasn’t buying it. He specifically targeted Amodei, claiming the administration had already taken steps to block potentially harmful actions from Anthropic. In the president’s framing, new regulations would primarily benefit China by slowing down American innovation.

Voluntary reviews, not binding rules

Earlier in 2026, Trump signed executive orders establishing voluntary review windows of up to 30 days for advanced AI models. The operative word there is “voluntary.”

The administration also imposed targeted export controls on specific Anthropic models. Mythos 5 and Fable 5 both faced restrictions due to cybersecurity concerns.

What the labs are pushing for goes considerably further. The proposals floating around industry circles include mechanisms like mandatory audit trails for frontier models, international collaboration frameworks, and structured accountability measures that would apply across the sector.

As of September 15, 2026, no new binding federal legislation on AI oversight has been enacted.

Why AI companies want to be regulated

For one, the leading labs have a competitive moat to protect. Binding safety standards tend to be expensive to comply with, which means they disproportionately burden smaller competitors and new entrants. OpenAI and Anthropic can absorb the cost of extensive safety testing. A scrappy startup building frontier models in a garage probably can’t.

There’s also the liability question. Without a regulatory framework, companies operating at the frontier of AI capability face enormous, undefined legal exposure. If something goes badly wrong with a model, the absence of government-mandated safety standards means companies have no compliance shield to point to.

And then there’s the genuine concern. Amodei has been among the most vocal AI leaders about existential risk. He co-founded Anthropic specifically to build AI systems with safety as a core priority. His essay calling for a global development slowdown reads as sincere, even if it also happens to align with his company’s competitive positioning.

The export controls on Anthropic’s Mythos 5 and Fable 5 models offer a preview of how the administration might continue to operate: targeted interventions on national security grounds rather than broad governance frameworks.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.