Cybersecurity stocks fall as new Anthropic tool sparks AI fears

Cybersecurity stocks declined for a second day on February 23, amid concerns over new AI security tools introduced by Anthropic.

The AI lab debuted a limited research preview of a service that scans software code for vulnerabilities and offers solutions on February 20.

Shares of companies such as CrowdStrike and Zscaler fell about 10%, while Netskope and Tenable dropped around 12%.

The iShares Cybersecurity ETF decreased by about 5%, reaching its lowest level since November 2023.

CrowdStrike CEO George Kurtz said that AI capabilities do not replace existing security platforms like Falcon, emphasizing the need for battle-tested solutions.

Several analysts noted that while AI could improve efficiency in specific tasks, it is unlikely to fully automate cybersecurity processes or replace comprehensive security platforms.

🔗 Source: CNBC

🧠 Food for thought

Implications, context, and why it matters.

Markets moved after the tool’s real-world results

Shares fell after details emerged about Claude Opus 4.6, which Anthropic said found more than 500 vulnerabilities in production open-source codebases during internal testing 1.
Some flaws had gone unnoticed for decades, even after expert review, since the model could reason through complex algorithms beyond the reach of traditional scanners 2.
Anthropic also tied the launch to misuse it observed, saying a state-sponsored group used its AI to automate 80-90% of a cyber espionage campaign, which raised the stakes for AI-based defenses 3.

The sell-off points to a platform shift, not a tool wipeout

The move fits a pattern where large tech platforms add security features, which can pressure specialist vendors by turning security into part of products customers already pay for 4.
Pattern-matching code scanner providers face the nearest risk, while declines in other names such as CrowdStrike tracked investor concern that AI will reshape security work rather than direct head-to-head competition 4.
Over time, the line between software development and security may fade as platforms that write code also review and repair it in the same workflow 4.
This could shift more security spending toward major AI platform providers that bundle protection, similar to how cloud providers grew into security leaders a decade ago 4.