Early in the year, a wave of cloud security and code security products launched, and investors had a scare. They lumped CrowdStrike (CRWD) and other cybersecurity names together as "AI losers," betting that AI would make the sector obsolete. The opposite happened. More AI adoption means more cybersecurity spending.
Cybersecurity works differently because these models have a dual use. If a model gets very good at catching cyber threats, it also gets very good at exploiting new attack methods. That two-way pressure keeps demand high. After the launch of Anthropic's Project Glasswing collaboration, the market came around to this view. Over the last few months, Palo Alto (PANW), Fortinet (FTNT), CrowdStrike (CRWD) and Okta (OKTA) all performed well. More growth lies ahead as companies work out how to secure their own use of AI and how to use AI to defend against new threats. Recent examples of the risk include Hugging Face and "agents against agents," where one agent goes rogue and searches the internet on its own.
CrowdStrike (CRWD)
The $460 figure often cited is a pre-split fair value estimate. CrowdStrike ran a 4-for-1 stock split on July 2nd, so on a current basis that number sits below street consensus. My base case is around $130, which is below where the stock trades now. I still like the company. From a structural and fundamental view, CrowdStrike, Palo Alto and other high-quality, wide-moat names should perform very well. The problem is that the market is pulling too much of that growth into the near term. The cyber story plays out over many quarters and many years, not in one blowout quarter.
My bull case for CrowdStrike sits slightly north of $200, so there is still some upside from current levels. That bull case depends on AI detection and response adding meaningfully to recurring revenue (ARR) and a real acceleration in AI-driven cyber demand. If CrowdStrike and Palo Alto report much better numbers than modeled in the next week or so, I would revise the fair value estimate upward. Both names carry bull cases that imply higher upside.
Okta (OKTA)
Okta reported strong earnings. As AI agents spread, companies need identity security for those agents, and the spending signals for securing agentic identity are clear. Okta is making a play in that space. I see it as roughly fairly valued, but the bull case, which assumes the AI drivers arrive sooner, shows more upside than where the stock trades now. Project Glasswing was a benefit to both Okta and CrowdStrike.
Top AI-spending beneficiaries
The three main beneficiaries of AI-driven cybersecurity spending are CrowdStrike (CRWD), Palo Alto (PANW) and Fortinet (FTNT).
Zscaler (ZS) is the most attractive on a price-to-fair-value basis. It trades well below fair value and stands to gain from higher AI spending, but only if it executes. Recent execution has fallen short, which is why the stock has lagged its cyber peers.
Palo Alto (PANW)
Palo Alto has the widest platform in cybersecurity, covering endpoint, network, cloud and SASE, and it is deeply embedded in enterprises. When companies raise their security budgets, going to an existing incumbent is the easy path. Execution under CEO Nikesh Arora has been strong. Palo Alto pushed platformization, or vendor consolidation, well before its rivals and doubled down on it. Its recent acquisition of CyberArc (CYBR) opened up the identity market it lacked before, which matters more now that identity is central to agentic AI and could drive solid growth. Palo Alto is also heavily weighted in the HACK ETF.
