AI-Powered Security Campaign Uncovers 5,000+ Software Issues in Bitcoin Projects

A coalition of security researchers and developers conducted a coordinated review of hundreds of Bitcoin-related open-source projects, revealing nearly 5,000 software issues in just over 30 hours. The team used artificial intelligence tools combined with human expertise to identify weaknesses.

The effort, which involved 16 researchers led by Calle, was supported by OpenSats, OpenCode, and AI inference sponsors. During the campaign, they reported 4,962 findings across 390 projects, with 85 critical issues and 635 high-severity findings.

The team used a unique approach that allowed human reviewers to guide AI systems during testing, helping uncover weaknesses that might have been missed by a single method. The campaign averaged about 166 reported findings every hour throughout the review.