The Coldcard vulnerability just cost the Bitcoin world over $100 million. Now a newly formed Bitcoin Red Team is hunting for the next disaster before it strikes.
Led by Calle, an Android developer behind Bitchat, and Rob Hamilton, CEO of Anchorwatch insurance, the team has already burned through $40,000 in AI tokens to scan 390 open-source repositories. They found 85 critical flaws and 635 high-severity issues in just 27.5 hours of auditing.
Calle's latest update reads like a war room report: 4,962 findings filed, 2.31 critical or high-severity issues per person per hour. The pace is relentless because the stakes are enormous.
The Red Team is using cutting-edge models, Kimi K3, GPT Sol, Fable, Opus and GLM5.2, among the priciest in the market. Access to OpenAI and Anthropic tools was initially bottlenecked, forcing the team to lean heavily on Chinese open-source models, a reality that stirred concerns about U.S. AI leadership in the Bitcoin space.
But momentum shifted fast. OpenAI granted access to GPT Sol. Hamilton's recent tweets hint that Anthropic doors opened too. OpenSats, a nonprofit 501c3, is footing the bill for the entire operation.
The Coldcard hack wasn't the only warning sign. Boltz exchange just paused operations to tackle AI-driven hacking attempts, a signal that the industry now sees automated attacks as an existential threat.
Coldcard users who haven't migrated their bitcoin to new seeds generated in secure firmware remain exposed. The window to act may still be open.
This article is informational and does not constitute financial or security advice. Always verify critical updates through official Bitcoin and hardware wallet channels.
