Every conversation about AI winners focuses on the companies building the models, selling the chips or renting the compute. Palo Alto Networks is none of those things. It is the company that secures all of them, and its quarterly results suggest that is the better business to be in.
Revenue hit $3bn for the first time, up 31% year on year and above both company guidance and Street estimates. Earnings per share of $0.85 beat consensus by 8%. Free cash flow came in at $910m against expectations of $735m. Remaining performance obligations, the backlog of contracted future revenue, reached $18.4bn, up 36%.
Wedbush raised its price target from $300 to $340, calling it a "Wembanyama-like quarter." The basketball analogy is unnecessary. The numbers speak for themselves.
AI is the accelerant
The connection between AI deployment and cybersecurity spending is direct. Every enterprise deploying AI agents, connecting models to internal data and running autonomous workflows is expanding its attack surface. More endpoints. More machine identities. More vectors for exploitation.
Palo Alto's management reported over 800 customer meetings in the quarter tied to AI security, including 240 focused on its Cortex platform and 150 customers interested in agentic endpoint security. The company noted that advances in frontier models, including Anthropic's Mythos, have increased the urgency for AI-native cybersecurity.
The $5m-plus account base grew 51% year on year. The $10m-plus base grew 49%. These are not small companies experimenting with security tools. These are large enterprises consolidating their security spending onto a single platform.
The 'platformisation' thesis
Palo Alto's strategy is consolidation. Rather than selling individual security products, it bundles network security, cloud security, identity protection and observability into a single platform. The company added roughly 110 net new platformisation deals in the quarter, including 20 from its CyberArk and Chronosphere acquisitions.
Next-generation security annual recurring revenue reached $8.13bn, up 60% year on year, with 65% coming from platformised customers. The acquisitions are tracking ahead of expectations.
The logic is that enterprises would rather buy security from one vendor with a unified platform than manage dozens of point solutions from different providers. Every new AI deployment reinforces that preference, because the complexity of securing AI workloads makes fragmented security architectures harder to maintain.
The valuation question
At $297, Palo Alto trades at roughly 79 times forward earnings. That is not cheap by any measure. CrowdStrike trades at a higher multiple, but Palo Alto's revenue base is more than double CrowdStrike's and its operating margins are comparable.
The company raised full-year revenue guidance to $11.42bn and earnings guidance to $3.77 to $3.79, both above consensus. It reaffirmed its target of 40% adjusted free cash flow margins by fiscal 2028.
The quiet trade
The AI investment narrative is dominated by chips, models and infrastructure. Cybersecurity rarely features in the conversation, which is odd given that every dollar spent on AI deployment creates incremental demand for security spending.
Palo Alto Networks is not building AI. It is protecting every company that is, and charging them more for it each quarter. In a market obsessed with finding the next AI winner, the most reliable one might be the company selling the locks for the doors everyone else is building.
