Synack CTO to Present Five Tests for Production-Ready AI Pentesting at Gartner Summit

LONDON, Sept. 17, 2026 /CourierPR/ -- Mark Kuhr, Chief Technology Officer and co-founder of Synack, will present at the Gartner Security & Risk Management Summit on September 24 in London.

According to Kuhr, AI pentesting systems often fail in production due to factors such as the need to handle authenticated workflows, custom business logic, and a dynamic attack surface. To address these issues, Kuhr will outline five key tests that security leaders should apply before trusting an AI pentesting system at scale. These tests include:

1. Orchestration: Ensuring the system can recover when a tool or attack path fails. 2. Exploit Verification: Proving that findings are truly exploitable rather than merely plausible. 3. Safety: Operating within scope without disrupting production systems. 4. Model Drift: Holding up as underlying models change. 5. Human Oversight: Allowing people to review, govern, and scale the system economically.

Attendees of the session will leave with a practical scorecard for evaluating AI pentesting systems, whether they are built in-house, purchased, or used in a hybrid approach. The five tests reflect the same standards Synack applies to its own platform, which combines agentic AI with a global community of security researchers.

Synack is the AI + human penetration testing platform for continuous security validation. Its Sara AI Pentesting system pairs agentic AI with a rigorously vetted global community of security researchers to expand testing coverage and prove real-world exploitability. The Synack Platform offers control and visibility across testing activity, validated findings, and remediation status. Founded by former NSA operatives, Synack has enabled nearly 10 million hours of security testing to protect critical assets across enterprise, public-sector, and highly regulated environments.