A volunteer security group discloses over a dozen critical flaws in Bitcoin projects after scanning 150 repositories using AI models.
A volunteer-led security initiative has identified more than a dozen vulnerabilities in Bitcoin software after scanning approximately 150 code repositories using advanced AI models. The group, dubbed a “Bitcoin red team,” focuses on probing wallets, cryptographic libraries, and infrastructure for weaknesses from an attacker’s perspective.
The effort has cost around $20,000 to date, funded privately, with no need for additional donations. The team employs models from OpenAI, Anthropic, and Z.ai to detect flaws and generate documentation, though developers note the process remains resource-intensive.
While the initiative aims to improve Bitcoin’s security, the findings highlight ongoing risks in blockchain infrastructure as adoption grows. No immediate market reaction was reported, but the disclosures could influence developer priorities and investor sentiment.