Jay Chaudhry sees artificial intelligence not as a threat to his company but as rocket fuel. The founder and chief executive of Zscaler Inc. has spent years telling investors and customers that the explosion of AI across enterprises will demand far more sophisticated protection. Billions in new revenue hang in the balance. So do new risks that few boards have fully grasped.
Chaudhry’s argument is straightforward. Enterprises already generate staggering volumes of data through their security platforms. Zscaler processes more than 500 billion transactions daily. Those feeds produce hundreds of trillions of signals every day. The company feeds this information into its models. The result powers better threat detection, risk scoring, automated copilots and, increasingly, agents that act on behalf of employees. Chaudhry calls the outcome “wonderful” AI.
But the same data that trains these systems has sparked questions. Cybersecurity manager Steven Swift posted on LinkedIn after Chaudhry’s remarks at a Cloud Security Alliance virtual summit. He noted that customers must enable deep packet inspection for Zscaler to see inside encrypted traffic. That inspection supplies the raw material for the company’s AI training. Privacy and trust concerns follow naturally. (SDxCentral)
Zscaler isn’t alone in chasing this opportunity. The company has acquired firms to broaden its reach. It bought SquareX to strengthen browser security suited to the AI era. It added Red Canary to integrate managed detection and response capabilities that feed findings back into models within minutes. And in recent months it announced plans to acquire Symmetry Systems to map and secure communications among AI agents themselves. Each move signals a conviction that agentic AI will create an entirely new attack surface.
Data from Zscaler’s own research backs the scale. Its 2026 AI Threat Report examined nearly one trillion AI and machine learning transactions generated by about 9,000 organizations during 2025. AI activity jumped 83 percent year over year. The number of applications driving those transactions quadrupled to more than 3,400. Data transfers to AI tools surged 93 percent to more than 18,000 terabytes. Finance and insurance produced the largest share of traffic. Technology and education saw the fastest growth. (Zscaler press release)
The vulnerabilities uncovered should worry any executive. Zscaler testers found critical flaws in 100 percent of enterprise AI systems they examined. Most could be compromised in just 16 minutes. In one case a system fell in a single second. Traditional defenses built for human-paced threats don’t stand a chance against machine-speed attacks. Deepen Desai, Zscaler’s executive vice president of cybersecurity, put it bluntly. “AI is no longer just a productivity tool but a primary vector for autonomous, machine-speed attacks by both crimeware and nation-state.”
Chaudhry has repeated a striking forecast in recent interviews. For every human employee, organizations will soon run 50 to 100 AI agents. Those agents will create identities, spawn sub-agents, access sensitive data and operate continuously. The perimeter isn’t users anymore. It’s autonomous workers that never sleep. At Zenith Live 2026 this week, Zscaler unveiled what it calls the industry’s first complete Zero Trust platform for agentic AI. New tools include an AI Broker, Endpoint AI Security and an AI Access Graph that maps identities, applications, agents and data flows. The company also expanded Project AI-Guardian with partners ranging from AWS and Google Cloud to OpenAI and system integrators such as Coforge and NTT DATA. (Yahoo Finance)
Investors have watched this evolution in the company’s financial results. Zscaler reported fiscal second-quarter 2026 revenue of $815.75 million, up 26 percent from a year earlier. Annual recurring revenue reached $3.36 billion. Full-year guidance rose. Chaudhry told CNBC that the need for cybersecurity has never been bigger and that the company’s Mythos platform is fueling growth. He has dismissed talk of an AI-driven “SaaS-pocalypse” as overblown. AI, he argues, cannot handle every deterministic task that matters in business. (CNBC)
Yet the stock has not always rewarded the optimism. Shares fell sharply after some earnings reports that beat expectations but failed to excite a market wary of software valuations. Chaudhry called one post-earnings drop a buying opportunity. The long-term case rests on the idea that AI multiplies both productivity and exposure. Every new agent becomes another identity that must be verified, every data transfer another potential leak, every autonomous action another vector for lateral movement.
But. The reliance on customer telemetry raises a deeper tension. Companies pay Zscaler precisely because they want their traffic inspected and protected. That same inspection supplies the training data that improves the platform for everyone. Chaudhry maintains this creates enormous value. Hundreds of trillions of signals don’t lie. They reveal patterns no human team could spot. They let models predict attacks before they fully form.
So the question facing CISOs isn’t whether to adopt AI. It’s how to govern the agents their organizations will inevitably deploy. Zscaler positions its Zero Trust Exchange as the answer. It claims to eliminate attack surfaces, block AI-specific threats, protect data everywhere, stop lateral movement and speed response. Legacy firewalls and virtual private networks, the company says, leave visibility gaps that grow wider as AI proliferates.
Chaudhry’s own background informs his confidence. The Indian-American entrepreneur has founded multiple security companies that were acquired before he launched Zscaler in 2007. He aimed then to build the Salesforce of cloud security. The shift to AI represents the next chapter in that vision. Where once the challenge was moving security to the cloud, today it is securing intelligence that moves at machine speed across global enterprises.
Recent moves suggest the bet is accelerating. Zscaler launched an AI and Cyber Threat Research Center with partners. It introduced AI Protect to address shadow AI risks. It continues to integrate acquisitions into a unified platform. Analysts debate whether these steps will translate into sustained market leadership against rivals such as Palo Alto Networks, CrowdStrike and Cloudflare. The data Zscaler controls gives it a defensible moat. The trust questions its CEO’s comments have surfaced could test customer willingness to feed even more information into the system.
One thing seems clear. The volume of AI activity is not slowing. Transactions measured in the hundreds of billions per year are becoming the baseline. Data flows measured in thousands of terabytes will only increase. Enterprises that treat AI security as an afterthought risk breaches that unfold faster than any incident response team can react. Those that build governance, visibility and Zero Trust controls from the start may turn the same technology into a competitive advantage.
Chaudhry doesn’t hedge on the outcome. AI agents are coming. The security demands they create will be measured in billions. Zscaler intends to capture its share.
WebProNews is an iEntry Publication