Application Security
,
Artificial Intelligence & Machine Learning
,
Events
Tenzai’s Pavel Gurvich on How Agentic AI Reshapes App Security and Testing Speed
AI-driven development accelerates application delivery, but it strains traditional security testing models, said Pavel Gurvich, co-founder and CEO, Tenzai. Faster code creation outpaces manual testing cycles. Security teams must balance release speed and risk exposure. Automated agents now fill this gap by testing applications at machine speed while maintaining coverage across evolving environments, Gurvich said.
See Also: How Technical Debt Puts Critical Infrastructure at Risk
Instead of relying on limited human expertise, organizations can use agentic AI to scale up advanced testing skills, with systems that not only evaluate source code but also deployment, configuration and integration layers. This broader visibility helps teams detect risks that static analysis often misses, he said.
“I would argue: You better have an agent of your own testing your applications before somebody else does,” he said.
In this video interview with Information Security Media Group at RSAC Conference 2026, Gurvich also discussed:
- How AI-generated code introduces new security gaps and missed edges cases;
- How agentic AI can add penetration testing expertise across teams;
- Continuous, parallel testing that replaces slow, sequential review cycles.
Gurvich leads Tenzai’s AI-native cybersecurity strategy, building autonomous AI capabilities to help enterprises secure code and strengthen resilience. Previously, he co-founded Guardicore, later acquired by Akamai, led Akamai’s enterprise security business, and previously served for 12 years in Israeli military cyber roles.

