As chip stocks plummet, "cybersecurity software" surges across the board, CrowdStrike and Palo Alto Networks hit new highs

AI safety warnings have triggered divergence within the technology sector, with cybersecurity ETFs outperforming semiconductor ETFs by more than 10 percentage points in a single day, marking the largest gap in history.

On Monday, CrowdStrike surged 13.8% to a record closing high; Palo Alto Networks jumped 13%, its largest single-day gain since April 2025; Fortinet closed up 9%. All three stocks ranked among the top gainers in the S&P 500 for the day. Previously, Anthropic CEO Dario Amodei warned about the rapid pace and potential loss of control in AI development. This statement put pressure on chip stocks but unexpectedly became a strong catalyst for the cybersecurity sector.

This trend has driven a historic divergence between the software and semiconductor sectors. According to Dow Jones market data, the iShares Expanded Tech-Software Sector ETF outperformed the iShares Semiconductor ETF by 10.67 percentage points in a single day, the largest one-day excess performance on record. Analysts believe the market signal is clear: as AI narratives face skepticism, capital is quickly rotating from computing hardware to security software.

AI Safety Warning Triggers Sector Divergence

In a blog post, Amodei detailed his concerns about AI "Recursive Self-Improvement" (RSI)—whereby once AI models possess the ability to self-train and iterate, they may experience uncontrollable leaps in capability.

He also mentioned a recent incident where an OpenAI AI agent attacked the open-source model platform Hugging Face, urging key industry players to slow the pace and prioritize safer AI progress. OpenAI CEO Sam Altman, as well as Tesla and SpaceX CEO Elon Musk, have both publicly expressed support for Amodei’s position.

This warning directly pressured chip stocks, which are highly sensitive to AI computing power demand, but the market quickly focused on another logic: regardless of the development pace of AI, security demand will only increase, not decrease.

The More AI Expands, the More Rigid Security Spending Becomes

According to Marketwatch, Evercore ISI analyst Kirk Materne noted in a Monday client report that regardless of the speed at which AI agents are adopted by enterprises, "these agents will still need to be protected, governed, and monitored," thus maintaining a "structural demand backdrop for cybersecurity and certain infrastructure names," even if the AI training pace slows.

Materne further stated, "A greater focus on AI safety may drive enterprises to pay more attention to identity management, data governance, observability, and security." He believes Okta, SailPoint, Palo Alto Networks, and CrowdStrike will be the first to benefit, as companies are prioritizing investments in identity security software.

He also specifically pointed out that cybersecurity vendors are accelerating collaboration with AI labs, particularly focusing on run-time security—that is, real-time protection of actively running systems.

Among all cybersecurity targets, Materne believes Palo Alto Networks and CrowdStrike "are best positioned to continuously capture growing security spending" because both companies have the differentiated advantage of "delivering integrated visibility and response capabilities across the entire security stack."

According to reports, Jefferies analyst Joseph Gallo stated in a report that while market momentum around AI agent security is building, it is still in its early stages overall. He expects to see "initial signals" later this year, while "substantial contributions" may not materialize until 2027 and beyond.

Gallo is also optimistic about identity security vendors as the first beneficiaries, citing "high visibility" demand in this segment and suggesting that the adoption of AI agents is accelerating the implementation of related products.

Notably, Gil Luria, Managing Director at D.A. Davidson, acknowledges the existence of AI safety risks but has a different view regarding the response path. In his report, he stated the right way for companies to address AI security threats is to "strengthen the code base"—that is, to improve the software’s own ability to resist attacks, rather than slowing down AI development.

Luria believes Microsoft, Amazon, and Google, with their vast customer bases, are "ideal candidates" to lead this direction. This view complements Materne’s and Gallo’s: regardless of which path the industry ultimately chooses, the core position of cybersecurity is hard to shake.