
Continuous AI pentesting that finds and fixes vulnerabilities
Report from 27 days ago
Veria Labs makes a security tool that plugs into code repositories and CI/CD systems to run continuous offensive-style checks as part of normal developer workflows. It runs on pull requests and builds, generates exploit proof‑of‑concepts for suspected issues, verifies them against staging to reduce false positives, and suggests patches developers can apply directly (verialabs.com).
The team also publishes offensive security research and responsible disclosures, indicating active capability beyond static scanning—for example, their write‑up on RCE vectors in developer AI tools and a related GitHub advisory that names a Veria researcher (blog post, GitHub advisory).
Top-down context:
Relevant 2024 markets total roughly USD ~12.1B: application security about USD 10.4B (IMARC) plus penetration testing about USD 1.7B (MarketsandMarkets). Both are growing at double‑digit rates.
Bottom-up calculation:
If priced at USD 15k–40k ACV per engineering org, adoption by 100k–250k CI/CD‑using orgs would imply USD ~1.5B–10B in addressable spend; this is a subset of the tens of millions of developers and their teams globally (JetBrains developer population).
Assumptions: