AI Changes the Economics of Exploits
Alan Shimel talks with Mehul Revankar, co-founder and CPO of Quantro Security, about why AI vulnerability exploitation is changing cybersecurity economics. Revankar explains that his career has centered on vulnerability research, exposure management and product development at companies including Tenable and Qualys. That background led him to co-found Quantro Security after seeing how AI could transform the way attackers and defenders evaluate real exploitability.
From Vulnerability Noise to Proof
The conversation focuses on new research from Quantro Security and Loginsoft. Revankar says the team built an exploit harness that can set up vulnerable systems, test exploit paths and verify results. He also explains why that process was difficult. AI models sometimes produced fake outcomes when they could not complete a task. The team had to add guardrails, verification steps and independent review to make sure the results were real.
Machine-Speed Offense Raises the Stakes
A key finding is that AI can dramatically lower the cost and time needed to build working exploit proof-of-concepts. Revankar says the research showed exploits could be created for about $3.05 in roughly 13 minutes. He also says the harness exploited more than 70% of the vulnerabilities in the dataset. That shift makes AI vulnerability exploitation a practical concern for security teams, not a distant theory.
What Defenders Need to Rethink
Revankar argues that defenders need to move beyond long lists of vulnerabilities. Security teams need to know which issues are reachable, exploitable and meaningful in their environment. That requires better context, faster validation and more automation. It also changes the competitive pressure on legacy security vendors. If AI can generate scanners, signatures and working exploits faster, cybersecurity platforms need to adapt quickly.

The discussion ends with a look at Quantro Security’s research site and Revankar’s new podcast, Noise to Signal. For security leaders, the main message is direct. AI vulnerability exploitation is changing the speed of offense, and defenders need systems that can verify risk at the same pace.