Palo Alto to enhance AI-driven security systems to tackle cyberattacks

Palo Alto to enhance AI-driven security systems to tackle cyberattacks

The firm seeks to take care of its clients’ business operations from rising hacking incidents

US company Palo Alto Networks has launched new versions of its cloud security platforms, Cortex Cloud and Prisma AIRS, integrating Protect AI's technology to secure AI applications.

It is also introducing agentic AI tools for cloud security, emphasizing human oversight in automated responses. The advancements will handle rising hacking incidents and infrastructure compromises that are impacting global businesses.

Its AI tools, combined with planned acquisition of Israeli peer CyberArk Software, are deepening its security systems amid a wave of high-profile cyberattacks that has hit global companies, including F5 and UnitedHealth Group.

Nikesh Arora, CEO of Palo Alto Networks, said some recent breaches displayed how back-end infrastructure compromises could expose thousands of customers by revealing vulnerabilities in shared source code.

The company said Prisma AIRS 2.0 would integrate technology from its recently acquired Seattle-based startup Protect AI, creating a combined platform to secure AI applications from development to deployment. It also uses AI systems to automatically find loopholes in other such systems.

Similarly, the Cortex Cloud 2.0 incorporates the agentic platform Cortex AgentiX, a cloud command center, and its unified view of cloud assets, to showcase risks and threats across cloud services of multiple providers.

The CEO added that customers would have the ability to tailor these agents for specific user roles. The pricing of the new agentic AI offerings would be consistent with the company's existing Cortex XSOAR platform that unifies and automates incident response across all security tools.

Arora further explained, "We're not going to take actions where the customers can't reverse them, or the customers can't have that a human in the middle type of activity. So, most of our agents will have humans in the middle.”

He added that agents were trained on 1.2 billion real-world security incident responses, and Palo Alto's standalone AgentiX platform was expected to launch early next year.