Y Combinator Co-Founder Says AI Labs May be 'Scared Enough' to Seek Government Regulation: 'Models Could be Dangerous'

Y Combinator co-founder Paul Graham suggested that AI labs seeking government regulation may be doing so because they have become increasingly concerned about the dangers and unpredictability of their own models.

AI Models Could Be Dangerous

On Sunday, Graham argued in a series of posts on X that skepticism toward AI labs’ calls for regulation may stem from underestimating the risks posed by increasingly capable models.

"The reason so many people look for an ulterior motive for the AI labs asking to be regulated is that they don’t grasp that models could be dangerous," he said.

He added, "But if you try assuming models are getting dangerous, or at least unpredictable, everything falls into place."

The reason so many people look for an ulterior motive for the AI labs asking to be regulated is that they don't grasp that models could be dangerous. But if you try assuming models are getting dangerous, or at least unpredictable, everything falls into place.

— Paul Graham (@paulg) September 20, 2026

AI Labs Face a Dilemma

Graham added that AI companies could face a dilemma if they believe development needs to slow down.

"If models are getting dangerous, you want to slow down," Graham wrote. "But you don’t want to slow down unilaterally and be left behind."

According to Graham, this could explain why companies seek government intervention to impose similar restrictions across the industry.

"You ask the government to make your competitors slow down too," he wrote, adding that companies cannot make that argument convincingly "without asking to be regulated yourself."

If models are getting dangerous, you want to slow down. But you don't want to slow down unilaterally and be left behind. So you ask the government to make your competitors slow down too. And you can't do that with a straight face without asking to be regulated yourself.

— Paul Graham (@paulg) September 20, 2026

‘Scared Enough’ to Seek Regulation

Graham dismissed the cartel explanation as preferable to what he sees as the more troubling possibility.

"I wish," he wrote regarding the cartel explanation. "But the real message is that the people making the models have become scared enough of them to take the otherwise unthinkable step of inviting the government in."

For me the message here is more alarming than a mere cartel. I wish. But the real message is that the people making the models have become scared enough of them to take the otherwise unthinkable step of inviting the government in.

— Paul Graham (@paulg) September 20, 2026

AI Revenue and Math Concerns

Last week, Brad Gerstner said leading AI labs needed to raise their combined annualized revenue from about $100 billion to at least $180 billion by year-end to sustain the AI trade and support massive infrastructure spending.

Meanwhile, twenty-five winners of the Fields Medal warned that AI companies’ race to solve mathematical problems could harm the field and raise plagiarism and attribution concerns, particularly when solutions were announced without sufficient time to verify methods or credit prior work.

Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.

Photo courtesy: Shutterstock