Palo Alto Networks (NASDAQ: PANW) announced an updated version of its Prisma Browser designed to secure autonomous AI agent interactions in the workplace. The cybersecurity company said the browser addresses security risks that emerge when employees use AI agents that act independently on their behalf.

The Santa Clara-based company stated that workers spend 85% of their workday in browsers, which have become central hubs for AI interactions. The updated Prisma Browser aims to protect against what Palo Alto Networks describes as new threats including shadow AI agents, prompt injection attacks, and agent hijacking.

The browser includes features that automatically discover user AI activity, enforce content boundaries to prevent data leakage to unmanaged AI tools, and block prompt injection attacks designed to manipulate AI agents. The system also distinguishes between human actions and automated AI tasks for compliance purposes.

"Organizations are unleashing a new workforce of agents, however, you cannot give autonomy without security," said Anand Oswal, Executive Vice President of AI & Network Security at Palo Alto Networks.

The browser enhancement is part of Palo Alto Networks' broader Prisma SASE solution, which the company describes as a comprehensive security platform. The solution includes autonomous operations capabilities designed to reduce manual IT troubleshooting and data protection features across AI tools and organizational systems.

Jonathan Jaffe, Chief Information Security Officer at Lemonade, commented on the browser's role in managing AI-related security threats, stating that "the browser will continue to be the single control point as agents end up doing things on behalf of the user."

The announcement was made through a press release statement from the company.