Capital One Open-Sources AI Security Tool VulnHunter

Capital One, one of the largest US financial institutions, operates in an industry where software flaws can expose sensitive customer data and trigger significant regulatory risks. As banks accelerate their use of generative AI and automated development tools, identifying exploitable weaknesses before code reaches production has become an increasingly important part of cybersecurity, compliance and technology-risk management.

Capital One has open-sourced VulnHunter, an agentic AI security tool designed to scan software source code while emulating how a hacker might look for weaknesses. The release gives outside developers and security teams access to the technology and could help them identify and remediate vulnerabilities earlier in the development cycle. The available event information did not specify an exact release date, investment amount or numerical performance benchmarks.