AI Is Coming for Cybersecurity and Investors Are Already Reacting

For years, cybersecurity has been one of the safest bets in tech, an industry built on the assumption that threats will only keep growing.

Now, investors are starting to ask a different question; what happens when AI becomes better at security than the companies built to provide it?

That question is suddenly hitting the market hard. Shares of major cybersecurity firms like CrowdStrike, Palo Alto Networks, and Zscaler dropped sharply after reports that Anthropic is testing a powerful new AI model, internally dubbed “Claude Mythos,” with advanced cybersecurity capabilities. 

The reaction wasn’t subtle. Some stocks fell between 4% and 8% in a single day, while broader cybersecurity ETFs also slipped as investors reassessed the future of the sector. 

What’s driving the panic isn’t just hype it’s the idea that AI could automate a large chunk of what cybersecurity firms actually do.

Modern cybersecurity companies generate billions by selling software and services that detect threats, scan vulnerabilities, and respond to attacks. But Anthropic’s latest AI efforts suggest those same functions could increasingly be handled by autonomous systems operating at machine speed.

In other words, the value could shift from human-driven platforms to AI-driven infrastructure.

That possibility is already showing up in early reports. The new model is said to include advanced capabilities in code analysis, reasoning, and security workflows areas that sit at the core of today’s cybersecurity offerings.

And unlike traditional tools, AI doesn’t just assist it can execute.

That’s the real disruption. If AI systems can continuously scan, detect, and respond to threats without human input, the need for layered security tools and the teams that manage them starts to shrink.

But the reality is more complicated.

Some analysts argue the market may be overreacting. Cybersecurity isn’t just about tools it’s about trust, compliance, and managing increasingly complex attack surfaces. And AI itself introduces new risks, including AI-powered cyberattacks that could make the threat landscape even more dangerous. 

That creates a paradox: AI could both reduce the need for traditional cybersecurity products and simultaneously increase demand for more advanced protection.

Still, the direction of travel is becoming clearer.

Instead of buying dozens of fragmented security tools, enterprises may begin to rely on a smaller number of AI-driven systems that can handle everything from detection to response. That shift alone could compress margins and redefine how cybersecurity companies compete.

For startups like Anthropic, it’s an opportunity to move up the stack—from being just an AI provider to becoming the intelligence layer that powers entire industries.

For incumbents, it’s a warning.

Because if AI becomes the default interface for security, then cybersecurity companies aren’t just fighting hackers anymore—they’re fighting the very technology that could replace them.