Bitcoin Red Team scans 150 codebases, discovers dozens of critical vulnerabilities

ChainCatcher reports that Bitcoin News posted on X that the volunteer organization Bitcoin Red Team has scanned 150 code repositories, privately disclosed over a dozen vulnerabilities, and identified critical issues during efforts to strengthen wallets, cryptographic libraries, and infrastructure following the COLDCARD vulnerability incident. This work has consumed approximately $20,000 in AI compute costs, covered by OpenSats. Researchers are using multiple AI tools, with many noting that open-source models such as Kimi K3 have become primary tools for security research, while leading proprietary models from OpenAI and Anthropic are considered more restricted in such analyses. The team plans to open-source its AI security tools to enable ongoing auditing of its own codebases. Researchers warn that Bitcoin may be the first industry to face a wave of AI-assisted vulnerability discovery, with similar efforts likely to expand to all open-source software next.