Zscaler Acquires Symmetry Systems to Map and Govern AI Agent Access Across the Enterprise

Zscaler (NASDAQ: ZS) announced its intent to acquire Symmetry Systems, a startup specializing in identity mapping and data access for AI security. The deal brings Symmetry’s access graph technology into Zscaler’s Zero Trust Exchange platform, creating what the companies describe as a foundation for governing how AI agents communicate with applications, data, and each other at enterprise scale.

Alongside the acquisition, Zscaler launched Project AI-Guardian, a strategic collaboration with six global system integrators: Cognizant, EY, HCLTech, Infosys, TCS, and Wipro. The program will use Zscaler’s AI Protect portfolio to help enterprises discover shadow AI, map data lineage across AI assets, and assess risks including supply chain vulnerabilities and sensitive data exposure.

The Access Graph

Symmetry Systems built its technology around an access graph that ingests enterprise-wide access logs from SaaS applications, public cloud services, data stores, and AI systems. The graph correlates those logs to show which identities are accessing which data and through which paths.

The core problem: traditional identity directories organize human users into stable groups mapped to applications and data. AI agents break that model. They operate with ephemeral identities and inherited permissions, according to Zscaler’s announcement, and their numbers are growing exponentially. The policies built for human users cannot scale to millions of autonomous agents communicating across interconnected systems.

“As enterprises rapidly adopt AI, the old playbook for governing access built around users and directories cannot scale to millions of AI agents,” said Jay Chaudhry, Chairman and CEO of Zscaler, in the company’s press release. “With Symmetry Systems, we are adding the access graph that maps how every identity, application, and data source connects across the enterprise.”

Combined with the Zero Trust Exchange, the access graph is designed to let security teams deploy AI agents with defined least-privilege policies, trace data lineage through chains of sub-agents and tools, detect anomalies in real time, and calculate blast radius if an agent or identity is compromised.

Project AI-Guardian

The GSI collaboration signals that agent security infrastructure is becoming a standard component of enterprise AI transformation projects, not an afterthought. Under Project AI-Guardian, partners will leverage Zscaler’s AI Asset Management, Secure Access to AI, and Secure AI Infrastructure tools to build discovery and risk mitigation functions for their clients.

As SecurityBrief reported, the strategy combines product expansion with a distribution model that relies on integrators already embedded in large enterprise technology programs. It also puts Zscaler in more direct competition with rivals over how businesses govern AI use, not just how they secure networks and user access.

Identity as the New Control Plane

Mohit Tiwari, CEO of Symmetry Systems, framed the deal in terms of a broader architectural shift. “We believe the dominant security platforms of the AI era will govern how information flows between identities across zero-trust networks,” Tiwari told Investing.com. “As AI disintermediates applications, endpoints, and traditional network boundaries, identities and data become the new control plane for enterprise security.”

The transaction is expected to close in the coming days, subject to customary closing conditions. Financial terms were not disclosed. Zscaler is scheduled to report earnings in the coming days, with its stock gaining 13.5% over the past week according to Investing.com.

The Agent Security Consolidation

Zscaler’s move arrives in the same 24-hour window as Xage Security’s launch of its own Zero Trust for AI platform. Two major cybersecurity vendors shipping agent governance products on the same day is not coincidence: it reflects a market that has moved past theoretical risk frameworks and into concrete infrastructure procurement. The question for CISOs is no longer whether to secure AI agents, but which vendor’s access model becomes the default.