Chipmakers Drop, Cybersecurity Surges on AI Slowdown Call — SkimNews
Get the Finance newsletter
Daily finance — markets, central banks, M&A, the prints that move money. Free.
- Nvidia fell about 3% and fellow chipmakers led a broad selloff, with Micron, Intel, Marvell Technology, and Applied Materials each dropping more than 4% and South Korean memory maker SK Hynix declining 7% in U.S. trading.
- Hewlett Packard Enterprise slid about 11%, Dell lost 6%, Oracle fell 4%, neocloud CoreWeave dropped roughly 7%, and liquid cooling provider Vertiv fell about 8% as the AI infrastructure complex unwound.
- Palo Alto Networks and CrowdStrike — both participants in early testing for unreleased OpenAI and Anthropic models — each jumped more than 13%, while Okta, Zscaler, Qualys, SentinelOne, and Netskope posted double-digit gains.
- Dario Amodei wrote in a Saturday essay that "we must slow the pace at which we improve the capabilities of AI models," but stopped short of pushing a complete halt; Sam Altman posted on X that "pacing" does "not mean 'stopping.'"
- The selloff was amplified by a viral social media firestorm after Anthropic researcher Jacob Coxon resigned saying the labs are "gambling with our lives," and colleague Evan Hubinger said there is a greater than 10% chance AI will "kill all humans" within the next decade.
- Quilter Cheviot's Ben Barringer countered the panic, noting that "inference is still the area that the industry is short in supply" and "demand still far outstrips supply," suggesting company revenues are unlikely to be hit even if training slows.
Why it matters: With over $1 trillion in annual AI infrastructure outlay now baked into U.S. earnings and profit margins, even a modest call for "pacing" — not a halt — frontier model development was enough to compress hardware valuations by 3-11% while rerating cybersecurity vendors like Palo Alto and CrowdStrike, who already participate in model testing, as the natural beneficiaries of AI risk mitigation.
Ask SkimNews
