Crypto
Home›Crypto›Regulation›Bitcoin red team says it found a dozen-plus vulnerabil…
Bitcoin red team says it found a dozen-plus vulnerabilities using AI
The volunteer effort said it has spent about $20,000 on AI scanning services and reported critical issues at a rate of roughly one per hour per person, while testing around 150 Bitcoin repositories.
A volunteer security initiative conducting what it calls a Bitcoin red team says it used frontier AI tools to scan roughly 150 Bitcoin repositories and make more than a dozen vulnerability disclosures. Decrypt reports that the effort is building an open-source AI platform aimed at auditing Bitcoin software, with developers saying it has uncovered critical weaknesses across areas including wallets, cryptographic libraries, and supporting infrastructure.
In a post on X, AnchorWatch CEO Rob Hamilton said the group has spent about $20,000 on AI services while developing and running the platform. Decrypt notes that Hamilton described the approach as drawing on multiple AI models, including Kimi K3 alongside OpenAI, Anthropic, and Z.ai models, and said the team received help from OpenAI for an additional scanning component.
The initiative also said it is averaging about one critical exploit per hour per person, according to a separate post on X from a pseudonymous Bitcoin developer identified only as Calle. Decrypt adds that Calle said the team reported critical vulnerabilities to several projects over a short period, and that the ongoing scanning work is expensive, with the group describing spending around $10,000 per day.
The effort did not name which specific projects were affected, nor did it provide technical details on the vulnerabilities. Decrypt said the announcement comes as AI is increasingly being used to locate security flaws across the crypto industry, including recent reported issues in other digital asset ecosystems.
Latest closeBitcoin $64,994.90 ▲0.2%