Enterprises are rapidly adopting AI agents, creating a governance gap that Zscaler aims to address with its new security platform updates. Recently, Zscaler announced enhancements to their Zero Trust Exchange platform, focusing on securing AI agents—a crucial step as enterprises deploy these technologies quicker than they can manage them effectively.
AI agents represent autonomous software programs capable of performing tasks, accessing sensitive data, and executing operations without direct human oversight. These capabilities have led to a swift integration across multiple sectors, amplified by the urgent need to control how these agents interact with enterprise systems and data.
At the core of Zscaler’s launch is a two-pronged approach to tackle security challenges in AI deployment. The AI Broker ensures secure communications for AI agents using protocols such as MCP and A2A. Meanwhile, the Endpoint AI Security component specifically targets threats arising from browsers, extensions, and local AI tools, directly on user devices.
A notable feature is the AI Access Graph. Originating from Zscaler’s acquisition of Symmetry Systems, it offers a detailed view of data and identity connections within an enterprise. This functionality helps security teams develop stringent access policies, minimizing unnecessary exposure and enhancing security postures.
Further expanding its capabilities, Zscaler has bolstered its AI Protect line with features covering AI asset management, access controls, and application protection. The enhancements aim to identify embedded AI across SaaS platforms and public cloud environments, offering visibility into endpoint AI activities. Security teams also benefit from improved controls, utilizing prompt extraction across numerous AI applications and red teaming for security assessment.
The current landscape underscores a critical issue: AI agent deployment is outpacing governance. Legacy security models designed around human users are now insufficient. AI agents function continuously, generate temporary identities, and make rapid access requests, all of which complicate traditional security systems.
Research highlights a burgeoning issue in security oversight. Adoption of AI agents is surging, yet governance frameworks lag behind. Without adequate oversight, agents often possess excessive permissions or operate without IT or security approval, introducing unknown risks into business environments. A significant 84% of senior leaders have identified unsanctioned agents as an emerging threat.
Toshiomi Takahashi, Senior Partner Business Manager at Zscaler, comments on the urgency of addressing these challenges. “With frontier AI advancing at an unprecedented pace, AI-related headlines have become part of our daily lives. This evolution brings both tremendous opportunities and significant risks, and the security landscape must evolve with it.”
Moving forward, Zscaler sets a precedent in addressing threats associated with AI deployments. The focus on data lineage, identity tracking, and agentic security emphasizes a shift in enterprise security investment. As organizations integrate AI agents into their operations, enhancing the infrastructure to manage and secure these agents becomes essential to safeguarding enterprise environments.
