Is the AI Bubble Bursting? - Newsweek

An industry-wide push to responsibly slow its advancement, and apocalyptic warnings from some of the technology’s leading developers, have dampened enthusiasm for artificial intelligence and raised concerns that one of its long-feared economic consequences—the AI bubble “bursting”—could be here.

Last week, Anthropic CEO Dario Amodei said unconstrained advancements risk rogue AI “swarms” capable of “taking over the entire internet” within a year, causing “catastrophic damage” that could total hundreds of billions of dollars. He proposed a plan, titled “pacing the frontier," which pushes for progress while ensuring adequate safeguards are put in place, and called on governments and fellow developers to pledge their commitment.

This came after former developers, notably former Anthropic researcher Jacob Coxon, warned that AI systems could pose an existential threat to the human race.

Coxon sounded the alarm on the technology again Monday night, telling Fox News that AI is moving “so incredibly quickly we won't have time to respond to it before it's too late.”

The former Anthropic researcher said international coordination would be needed to manage the risks posed by AI. Asked what gave him hope that China would cooperate with the U.S., Coxon said he was “not an expert” on geopolitics but that the alternative could be catastrophic.

“And what’s needed is essentially a deal, like a big change, and I think President Trump could be a great person to make a deal in the immediate future—like a good solid deal with China,” Coxon told Fox.

In an interview with CNN’s Anderson Cooper that aired Monday night, Amodei was asked whether he agreed with Coxon’s post on X, which said the people building AI “earnestly believe that it could kill us all by the end of the decade.”

“Look, I’m still worried about all of these things. I stand by these concerns about labor displacement if the progress of the field is too fast and if society doesn't respond in the right way,” Amodei said.

President Donald Trump called claims that AI will take over the world a “hoax.”

“I am the Hoax Buster, and I’m right now breaking another Hoax — That AI is going to take over, consume, and destroy the World, and that Robots will be marching into our Cities, and getting rid of us all!” Trump posted on Truth Social.

Trump AI adviser David Sacks also questioned why Amodei would continue running a frontier AI company if he believed the technology could ultimately endanger humanity.

“How are you gonna make this safe, and if you don’t think you can—you should probably either shut down the company or step aside and let someone else lead it,” Sacks told CBS News.

Amodei said he remains concerned about the risks posed by artificial intelligence, including the possibility of catastrophic consequences if the technology advances too quickly.

“It’s not that the things we’re building are bad, it’s that it’s happening so fast that if we don’t slow down a little bit we’re going to make a mistake,” Amodei said. “And so we need to slow down and we need to take the proper time to do this right.”

Amodei said Anthropic has always viewed AI as a powerful technology with “a lot of benefits” but also “very serious risks.”

Prominent tech figures, including Elon Musk and former Google DeepMind CEO Demis Hassabis, have since echoed Amodei's calls, as has OpenAI CEO Sam Altman, who said those pioneering the tech “could lose control of the future to AI.”

Amid these warnings, AI-related equities have come under pressure across markets in Asia, Europe and the United States, linked to fears that a coming development slowdown could further postpone the so-far illusory returns on the staggering investments made in the technology.

Broadly, the AI “bubble” has come to describe the potential disconnect between the enthusiasm surrounding AI and the technology’s ability to deliver the revolutionary profitability investors expect. Like the Dot-Com bubble of the late 1990s, experts have said the sizable gains made by AI-linked stocks are not justified by current revenues or business models, evidence of what former Federal Reserve Chairman Alan Greenspan dubbed “irrational exuberance” in the market.

AI-related firms have become a significant driver of market gains in recent years, while their associated infrastructure plays an increasingly central role in the U.S. economy, meaning that a sudden correction or “bursting” of this bubble could have economy-wide consequences.

And while leading voices in the AI space have long advocated for responsible advancement, the consensus that appears to be forming is one of responsible development that experts say may have spooked investors.

“The warnings come as AI systems become more capable of acting independently, including in areas such as cybersecurity,” said Benjamin Arold, an economics professor at Cambridge University. “Developers are concerned that their ability to build more powerful systems may be advancing faster than their ability to keep them under control.”

“The warnings may have contributed to today’s selloff,” he told Newsweek. “If development slows or safety requirements become more demanding, investors may have to wait longer for returns on the enormous sums being spent on AI.”

And some of this fear is not speculative. In an interview with Fortune, published Saturday, Altman said OpenAI would be pushing back its plans to go public until at least 2027.

“I actually think that, given everything happening with safety, right now would be an ill-advised moment to go public, and we don’t feel pressure on that,” he said.

These factors appear to have taken some wind out of the sails of AI-linked equities in Asia, Europe and the U.S. on Monday.

As of noon ET, chipmaker Nvidia had fallen about 3 percent, with Intel down more than 5 percent, while in Tokyo, SoftBank—a major stakeholder in OpenAI—had fallen 10.7 percent by the closing bell.

But despite the heavy losses, economists and scholars of past financial “bubbles” say the selloff does not yet signal the mass-market correction many fear.

William Quinn, a financial historian and co-author of Boom and Bust: A Global History of Financial Bubbles, told Newsweek that the current market panic could also be fueled by the possibility of rising global interest rates. Quinn said that a decline in AI-linked stock prices, while rare, is not the “same thing as a bubble bursting,” and that these have for the most part risen in tandem with actual earnings.

“Prices broadly reflect optimism that those high earnings can be sustained. The market could be wrong about that,” he said, “but it's not the type of thinking that we usually associate with a bubble.”

“One bad morning does not mean the AI bubble has burst,” said Arold. However, he added that this, if sustained, "could mark the start of a broader correction in A.I. stock prices” if the high valuations and investments currently enjoyed by AI firms fail to “translate into equally enormous profits.”

“If revenues and productivity gains fall short of expectations while those costs rise, stock prices could have much further to fall, even if AI itself proves transformative,” he told Newsweek.

Contact Newsweek editors on this story: Ben Kelly, Sam Wilson, Gray R. Thomas