TradingKey - On September 14 ET, cybersecurity stocks broadly advanced as investor concerns grew that artificial intelligence could pose risks to humanity, and several industry leaders publicly called for slowing down the pace of model development.

As of press time, CrowdStrike (CRWD) rose 15.25%, Palo Alto Networks (PANW) rose 13.98%, Okta (OKTA) rose 12.82%, Fortinet (FTNT) rose 8.87%, Cloudflare (NET) rose 7.79%.

The trigger for this rally was collective statements from AI industry leaders regarding the risks of artificial intelligence. Investors worry that AI could harm humanity, leading some industry insiders to begin calling for a slower pace of development.

CrowdStrike stock price chart, Source: TradingView

Just over the weekend, Anthropic CEO Dario Amodei warned that the AI industry needs to slow down model development to avoid potentially catastrophic consequences. Subsequently, OpenAI CEO Sam Altman stated that the development of frontier AI needs to proceed "at a measured pace," but should not stop. SpaceX (SPCX) CEO Elon Musk, who rarely agrees with Amodei and Altman on these issues, also remarked that the industry will have to slow down a bit. Microsoft (MSFT) CEO Satya Nadella similarly joined the discussion, stating that AI development requires "deliberate pacing."

The market's interpretation is that enterprises will consequently increase investment in infrastructure security, which could drive faster revenue growth for cybersecurity companies.

Notably, the strength in the cybersecurity sector stands in sharp contrast to the weakness in AI hardware. The Philadelphia Semiconductor Index is currently down over 5%, with Micron Technology (MU) and Marvell Technology (MRVL) falling over 5%, while Intel (INTC) and AMD (AMD) dropped over 4%.

The core thesis supporting this logic comes from Evercore ISI analyst Kirk Materne. In a note to clients, he stated that demand for cybersecurity and infrastructure should be "largely orthogonal" to the pace of frontier model advancement.

Materne's reasoning is that whether autonomous agents enter enterprises faster or slower than currently expected, these agents must be protected, governed, and observed; they also require controlled access to critical enterprise data. Therefore, regardless of the progress in frontier model training, the structural demand backdrop for cybersecurity and select infrastructure assets should remain relatively intact. This view redefines cybersecurity from an "AI beneficiary stock" to an "AI prerequisite stock." The former relies on the realization of upside in model capabilities, whereas the latter merely relies on AI actually being deployed.

Materne believes that the rationale for Okta and SailPoint benefiting lies in identity. Because identity remains a "top priority" for enterprise customers, spending related to identity security exhibits strong stickiness. Meanwhile, the additional advantage for Palo Alto Networks and CrowdStrike stems from their positioning as "platform vendors": they can provide consolidated visibility and response capabilities across a broader security stack, making them, in Materne's view, the most likely to capture this spending over the long term.

Find out more