Palo Alto Networks Says AI Security Needs Machine-Speed Controls

Palo Alto Networks Chief Information Officer Meerah Rajavel says cybersecurity must operate at the speed of artificial intelligence as companies deploy more autonomous agents. The statement is partly a product pitch, but it identifies a real operational mismatch: automated systems can act faster than traditional approval and response processes.

AI agents can search data, call software tools and make changes across corporate systems. That creates productivity gains, but it also turns an agent into a new kind of identity with permissions that attackers may steal or misuse. Security teams need to know which agent acted, what credentials it used, what data it reached and whether its behavior remained within policy.

Rajavel emphasized AI-powered defenses, governance, guardrails and continued human oversight. Those controls are complementary. Automated detection can respond to machine-speed attacks, while approval gates remain useful for high-impact actions such as moving money, changing production infrastructure or exposing sensitive data. Least-privilege access and short-lived credentials can limit damage when an agent behaves unexpectedly.

Palo Alto Networks has a commercial reason to frame the problem this way. Its platform strategy seeks to consolidate network, cloud and security-operations tools, and AI creates another reason for customers to integrate data across those products. The company's fiscal 2026 results showed continued growth, but management also lists intense competition, customer consolidation, cloud-provider security products and the risks of AI agents as material uncertainties.

The investment case therefore depends on more than the size of the threat. Palo Alto must convert AI security demand into durable subscription growth without sacrificing margins or relying on acquisitions that are difficult to integrate. Customers may prefer one broad platform, but they can also resist lock-in or keep specialized vendors for critical functions.

The interview did not announce a new contract, product or financial target, so it should not be treated as a direct earnings catalyst. Its value is strategic: enterprise AI adoption expands the attack surface and makes identity, telemetry and response automation more important.

Investors should look for evidence in next-generation security ARR, platform adoption, retention and large customer wins. The faster AI acts, the more valuable effective controls become, but only if those controls are measurable, auditable and safe.