Palo Alto Networks shows how AI is changing cybersecurity spending.

Source Domain: startupfortune.com

AI is turning cybersecurity from a defensive line item into a platform decision, and Palo Alto Networks just gave investors a clear look at that shift.

Palo Alto Networks did not simply report a stronger quarter. It showed how quickly AI is rewriting the security budget. The company said on June 2 that fiscal third-quarter revenue rose 31% from a year earlier to $3.0 billion, with customers moving faster to secure AI deployments that are now spreading across cloud systems, identity tools and software teams.

That matters because AI spending is often discussed as if it belongs only to chipmakers, cloud providers and model builders. The more practical story is wider. Once companies start using AI agents, coding assistants, automated workflows and internal models, they also create new access points, new machine identities and new ways for data to leak. Security spending follows that risk.

According to Palo Alto Networks’ fiscal third-quarter release, Next-Generation Security annual recurring revenue grew 60% year over year to $8.1 billion, while remaining performance obligations rose 36% to $18.4 billion. Those numbers are important because they point beyond one good quarter. They suggest customers are committing to longer-term security contracts while they rethink how much of their infrastructure should sit with one provider.

The strongest part of Palo Alto’s story is not just demand. It is consolidation. The company has spent heavily to make itself harder to replace, adding CyberArk for identity security and Chronosphere for cloud-native observability. In the latest quarter, those two businesses contributed $388 million of revenue, $1.6 billion of Next-Generation Security ARR and $1.8 billion of remaining performance obligations.

That is a large contribution, but the strategic point is larger. AI security is not one product. It touches who can access systems, what software agents are allowed to do, how cloud workloads behave, how incidents are detected…

Source