Palo Alto Networks CEO Nikesh Arora expects $5 trillion in capital spending on AI infrastructure over the next five years, and he says that buildout will require an entirely new security stack. The company also reported fiscal fourth-quarter revenue that beat analyst estimates, with strong growth in its next-generation security business.
The $5 Trillion AI Buildout
Arora laid out the scale of the coming AI infrastructure wave during the company's earnings call. He said companies must build a completely new security stack to protect the AI infrastructure they're putting in place. That's not a tweak to existing tools, he argued, but a fundamental rethinking of how security works when AI systems are the target.
The $5 trillion figure covers capital spending across the industry over the next five years. Arora didn't break down where that money goes, but the implication is clear: a lot of it will need to be secured, and the old playbook won't hold up.
The $1 Trillion Security Debt
Arora also put a number on what he calls global cybersecurity debt: roughly $1 trillion. He arrived at that figure by looking at how long security equipment typically lasts — about seven years — and multiplying that by annual spending of $200 billion to $300 billion. The result is a massive backlog of outdated systems that need to be modernized to defend against automated threats.
That debt isn't just a cost problem. It's a timing problem. As AI makes attacks faster and more sophisticated, the gap between what's deployed and what's needed widens. Arora's point is that the industry can't wait another seven years to catch up.
Anthropic's Mythos Model
Arora highlighted Anthropic's Mythos model as a major shift in cybersecurity. Unlike previous AI models that mostly helped defenders, Mythos can identify and exploit software vulnerabilities on its own. That changes the threat landscape in a fundamental way, he said, because attackers now have a tool that can find holes faster than human researchers can patch them.
For security vendors, that means the bar for what counts as a defense is rising. A system that was good enough last year may not be good enough next year, and Arora's comments suggest Palo Alto Networks is positioning itself to sell that upgrade.
Strong Q4 Numbers
The earnings themselves gave investors plenty to chew on. Palo Alto Networks reported fiscal fourth-quarter revenue of $3.41 billion, up 34% year over year and ahead of the $3.35 billion analysts had expected. Adjusted earnings per share came in at $1.02, four cents above the consensus estimate.
Next-generation Security annual recurring revenue hit $9.10 billion, a 63% jump. Remaining performance obligations rose 34% to $21.2 billion, a sign that customers are signing longer and larger deals.
For fiscal 2027, the company guided to $14.10 billion to $14.20 billion in revenue, ahead of the $13.79 billion analysts had forecast. That guidance suggests Arora sees the AI security push as a growth driver, not just a defensive play.
The question now is whether the rest of the industry can move fast enough to close that $1 trillion gap before the automated threats Arora is worried about become the norm.
