AI alarm bells are ringing

After years of seemingly endless hype, the AI bubble may finally be starting to pop—thanks to the very people who inflated it in the first place.

More and more AI insiders are warning that the technology may be advancing too fast. Anthropic researcher Jacob Coxon sounded the alarm last week when he resigned, declaring that people building frontier AI believe it “could kill us all by the end of the decade.” Then over the weekend, Anthropic CEO Dario Amodei called on the industry to “pace the frontier,” warning that “in 6–12 months such a swarm could be capable of taking over the entire internet.” Sam Altman and Elon Musk were among the major AI figures that backed the push for more caution.

Hold your robot horses

Not everyone is buying the doomsday scenario:

• Nvidia CEO Jensen Huang argued that some of the recent safety fears are being overhyped by companies that stand to sell the solutions. “What better way to create demand than to create a problem?” Huang said.
• President Trump rejected calls to slow development, arguing that the US can’t afford to give up its lead over China. “WHOEVER WINS AI, WINS!” Trump wrote on Truth Social.
• China also dismissed the push for a slowdown, with a spokesperson for China’s Foreign Ministry warning that, “Fear-mongering, confrontation, competition will just disrupt [the] process of global AI governance.”

Still, the unusual agreement among rivals like Amodei, Altman, and Musk was enough to get Wall Street’s attention, especially because the AI boom now extends far beyond ChatGPT into an ecosystem spanning semiconductors, infrastructure, and energy.

A new era begins

The market’s reaction today illustrated which companies will feel the most pain from an AI slowdown—and which stand to benefit.

The losers: Companies most exposed to continued AI expansion took the biggest hit. The Philadelphia Semiconductor Index fell 5.86%, while Nvidia, AMD, Micron, and other chip names dropped. OpenAI has scrapped plans for a 2026 IPO, while Anthropic’s own path to market has become murkier.

That’s bad news for a market that has become increasingly dependent on AI. Four of the biggest AI spenders make up roughly 17% of the S&P 500, while the largest hyperscalers are on track to drive more than half of all S&P 500 capital spending by 2028.

The winners: Companies that have been threatened by rapid AI progress moved the other way. Software names including ServiceNow, Adobe, and Salesforce rose as investors bet that slower development could reduce the threat of AI replacing traditional software. Cybersecurity stocks like CrowdStrike and Palo Alto Networks rallied sharply, while Alphabet, Microsoft, and Meta gained, as investors considered whether a slowdown could give the two tech giants’ lagging models more time to catch up to the competition.

Zoom out: The AI trade was built on one simple assumption: Progress will keep accelerating. This weekend’s warnings challenged that premise, forcing investors to consider what happens if the race slows before the spending does. For now, though, they’re left with nothing but a big question mark hanging over the market’s most important trade.—SY