The Bitcoin Red Team is using Chinese AI models to scan the entire Bitcoin open-source ecosystem for security flaws. In August, the group filed 4,962 findings across 390 projects, including 85 rated critical and 635 rated high severity. Developers have confirmed numerous real critical and high vulnerabilities, particularly in Lightning software described as « more broken than the average. » Facing restrictions from U.S. models by OpenAI and Anthropic, the team relies on Kimi K3 from Chinese startup Moonshot AI and GLM 5.2 from Z.ai to conduct its audits.
Source: Read the original article
