

Using AI Agents to Solve Cloud Vulnerability Overload
14 snips Jun 17, 2025
In this engaging conversation, Harry Wetherald, Co-Founder & CEO of Maze and former product manager at Tessian, dives into the complexities of modern vulnerability management. He reveals how traditional tools often yield 90% false positives and advocates for AI agents that emulate security engineers to pinpoint real threats. Harry also discusses the hurdles of developing effective AI solutions and how they can foster collaboration between security and engineering teams, ultimately enhancing cloud security practices.
AI Snips
Chapters
Transcript
Episode notes
Vulnerability Overload Crisis
- Vulnerability overload stems from a massive increase in vulnerabilities and a sharp decrease in time to exploit them.
- Security teams face growing challenges with unchanged resources, heightening the risk landscape.
Cloud Vulnerability False Positives
- 90% of cloud vulnerability findings by tools like CNAPPs are false positives, creating noise for security teams.
- False positives arise from logical scanner errors beyond just reachability issues.
Understanding The Wall of Red
- The "wall of red" vulnerability alerts feels overwhelming because most are false alarms.
- Initial CNAPP deployment lights up many findings but leaves teams stuck managing noise.