Chip stocks tumble on the AI slowdown calls. $1 trillion a year of spend just became a market variable

Wall Street repriced the AI supply chain on Monday, on a weekend argument rather than a product.

Micron, Intel, Marvell Technology, and Applied Materials each dropped more than 4 percent. Nvidia fell about 3 percent. SK Hynix lost 7 percent in US trading, CNBC reported.

Server-side names followed. Hewlett Packard Enterprise slid about 11 percent, Dell lost 6 percent, Oracle about 4, and neocloud CoreWeave about 7.

The trigger came from the weekend, not from a filing. Anthropic chief executive Dario Amodei asked frontier labs on Saturday to slow the pace of capability gains, and OpenAI's Sam Altman plus SpaceX's Elon Musk publicly agreed.

"They may be really good at models, but they're not good at talking stocks, and what they're doing is freaking the market," Melius Research's Ben Reitzes said on Monday, per CNBC.

Altman stated his own limits in a post that same day.

"Progress has been rapid and will continue to be. But it should be slower than it otherwise could be; interventions like safety cases and monitoring have significant costs," Altman said.

What slowed, what gained

The split is the analytical carry of the day. Damped model training slows some demand, but inference demand "still far outstrips supply," Quilter Cheviot analyst Ben Barringer said, so revenue holds even at a softer pace.

Cybersecurity stocks read the same worries in reverse. Palo Alto Networks and CrowdStrike jumped more than 13 percent each, with the expectation that firms will spend more on security as models get more capable, CNBC reported.

Boockvar framing is the money line: "The U.S. economy, the U.S. stock market, the trajectory of U.S. earnings and U.S. profit margins are all in on this capex spend," Peter Boockvar said, so anything that touches the buildout ripples everywhere.

Precedent says a single speech does not reprice compute. What changed here: safety rhetoric moved from a research note to a funding condition, and lenders and boards who price AI projects will now ask about pace.

The unresolved edge is whether the market treats this as a one-day move or a template. Monday's losses recede or compound based on what the labs actually announce.

I get the sense the supply chain now carries two prices. One for compute, one for conflict about how fast it should run.

Source