Palo Alto Networks today, during the virtual Ignite conference, added a raft of capabilities to both protect artificial intelligence (AI) applications and leverage AI to automate security workflows.

In terms of protecting AI applications, Palo Alto Networks has updated its Prisma AIRS platform for securing artificial intelligence (AI) applications to include capabilities the company gained via the acquisition of Protect AI earlier this year.

Anand Oswal, executive vice president of network security for Palo Alto Networks, said Prisma AIRS 2.0 enables cybersecurity teams to extend the reach of the core Palo Alto Networks platform to now secure the entire lifecycle of an AI application or agent. That capability is critical because the pace at which these applications are being built and deployed is rapidly exceeding the ability of cybersecurity teams to keep pace, he added.

Additionally, Palo Alto Networks is launching Cortex AgentiX, a security orchestration, automation and response (SOAR) platform for securely building, deploying and governing AI agents with cybersecurity workflows. At the same time, Palo Alto Networks in the first half of 2026 will be updating Cortex Cloud to add AI agents and a revamped Cloud Command Center to a platform that combines a cloud-native application protection platform (CNAPP) with a cloud detection and response (CDR) platform.

The Cloud Command Center now also provides tighter integration with an Application Security Posture Management (ASPM) module that Palo Alto Networks added to its portfolio earlier this year.

In general, Palo Alto Networks continues to make a case for standardizing cybersecurity on a set of integrated platforms provided by a single vendor. For example, instead of acquiring a separate platform from another vendor, Prisma AIRS 2.0 provides a simpler approach for securing AI applications inline as they are being built and deployed in a way that spans everything from inspecting the underlying model to protecting AI agents in runtime as they are deployed, said Oswal.

Other Prisma AIRS 2.0 capabilities also include the ability to discover and inventory every AI agent in use, regardless of whether it has been sanctioned by IT and cybersecurity teams, red team testing tools for discovering vulnerabilities, and scanning tools for discovering threats to AI models that involve architectural backdoors, data poisoning, and malicious code hidden within the model layers.

While Palo Alto Networks is moving to secure AI applications, it is also embracing AI to help reduce toil. Cortex Cloud 2.0, for example, makes it simpler to use AI agents to aggregate and enrich threat intelligence, automate responses to email threats, investigate endpoint issues and apply network security policies.

Additionally, cybersecurity teams can also use a no-code tool to build their own AI agents using more than 1,000 pre-built integrations provided by Palo Alto Networks.

The AgentiX SOAR platform, meanwhile, extends that capability by using a set of more sophisticated pre-built AI agents that are able to dynamically plan, reason, and execute workflows. Trained by Palo Alto Networks using 1.2 billion playbook executions, AgentiX also comes with over 1,000 prebuilt integrations and native support for the Model Context Protocol (MCP) developed by Anthropic to provide AI agents with access to legacy data.

It’s not clear at what pace cybersecurity teams are embracing AI to automate workflows, but the rise of thousands of AI agents across the enterprise that all have a unique set of capabilities is likely to force the issue. The only thing that remains to be seen now is not so much whether cybersecurity teams will be employing AI but rather the degree and extent to which AI agents can be relied on to autonomously help defend an attack surface that only continues to expand further with each passing day.