WASHINGTON-Chief executives aren't buying President Trump's argument that AI's threats to humanity are a hoax.

At an invitation-only gathering of dozens of America's top executives in Washington this week, business leaders overwhelmingly disagreed with the president's assessment that dangers posed by artificial intelligence are being exaggerated. In a flash poll of attendees at the Yale School of Management event, 93% said Trump was incorrect in calling the technology's potential catastrophic dangers a hoax, as he did earlier this week.

"It was stunning," said Jeffrey Sonnenfeld, a Yale management professor who organized the CEO meeting. In conversations with executives both in the room and in debriefings afterward, Sonnenfeld said business leaders told him they wanted more discussions on ways to safeguard against AI and its possibilities for harm. "The sentiment was: It's real."

The forum was a chance for business leaders to compare notes ahead of the White House's state dinner with China's leader Xi Jinping next week, where the global race to build the most advanced AI is likely to loom large. Trump has dismissed calls for tighter federal regulations, arguing that such guardrails would financially ruin the industry and give China an edge. Leaders of the biggest AI companies, including Anthropic and OpenAI, said over the weekend that the industry needed to slow development of advanced AI to ensure its safety.

The business leaders hailed mostly from industries and parts of the country outside Silicon Valley and included CEOs from the toolmaker Snap-on, Motorola Solutions and glassmaker Corning, who received an award. The room was a mix of Republican and Democratic executives and government officials, including former Vice President Mike Pence and former House Speaker Nancy Pelosi.

Many didn't want to speak publicly for fear of alienating the Trump administration, but a series of poll questions let them share their concerns collectively. Asked whether Trump should address the need to collaborate with China on AI-safety guidelines during his meeting, 88% of them said yes.

As to whether Trump would actually do so, the executives expressed less confidence: About three-quarters of respondents said they didn't expect the president to push for joint guardrails.

The White House didn't immediately respond to a request for comment.

Chris Layden, the CEO of staffing and recruiting firm Kelly, who attended the Yale event, said both AI's potential for harm and the risk of losing the AI race to China need to be addressed. He said he also wants to see government and business focus more on what the technology means for jobs and U.S. workers.

"We needed a U.S. workforce strategy on AI readiness yesterday," Layden said. "We've got to focus on the trust that the American worker has with AI, and the opportunity it has to help make us better learners, more efficient, more productive."