Zscaler announced the acquisition of Symmetry Systems on May 21, 2026 [2], to extend its security capabilities for autonomous AI agents [2].
The move comes as corporate networks face a massive influx of automated tools that can bypass traditional security perimeters. If these agents are not secured, they could create unprecedented vulnerabilities in corporate data infrastructure.
Jay Chaudhry, founder and CEO of Zscaler, said the company expects to see billions of autonomous AI agents on corporate networks [1]. He said these agents will require robust cybersecurity protection to operate safely [1].
To address this, Zscaler is integrating Symmetry Systems' technology into its existing framework. A company spokesperson said the acquisition will enhance the ability to map and secure AI-agent communication across the Zero Trust fabric [3]. The announcement originated in San Jose, California [3].
Beyond the security of agents, Chaudhry addressed the physical limitations of AI growth. He said a current chip crunch is putting pressure on memory costs and overall AI infrastructure [4]. This hardware shortage creates financial and operational pressure on companies scaling their AI capabilities.
Chaudhry said the company must innovate to stay ahead of these infrastructure costs [4]. The strategy involves combining Zero Trust principles with access graph technology to better monitor how AI agents communicate, and interact within a network [3].
“"We will see billions of autonomous AI agents on corporate networks, and they will need robust cybersecurity protection."”
The shift toward autonomous AI agents represents a transition from AI as a tool used by humans to AI as an independent actor on a network. By acquiring Symmetry Systems, Zscaler is attempting to solve the 'visibility gap' where traditional security tools cannot track the rapid, automated communications between AI agents. This reflects a broader industry trend where cybersecurity must evolve from protecting static user accounts to managing dynamic, machine-to-machine identities.
