Bitcoin Red Team AI Audit Flags Hundreds of Critical Vulnerabilities

A volunteer group called the Bitcoin Red Team has conducted a large‑scale security audit of Bitcoin‑related software using artificial‑intelligence tools. In the first 55 hours the team, comprising 24 members worldwide, examined 425 projects and generated 6,700 findings, of which 1,029 were classified as high or critical. Earlier results after 27.5 hours covered 390 projects with 4,962 findings, including 85 critical and 635 high‑severity issues.

The audit leveraged multiple AI models to scan wallets, cryptographic libraries and infrastructure code, with human experts guiding the analysis and confirming the most severe reports. Team members note that the main challenge now is triaging and delivering the reports to project maintainers, as not all flagged items have been verified as real vulnerabilities. Calle, a developer of the Cashu protocol, described the situation as “extremely serious.” The effort highlights both the growing capability of AI to uncover software flaws and the parallel risk that the same technology could be used by attackers to discover zero‑day exploits.