A Bitcoin security resource group said it used advanced AI models to inspect about 150 Bitcoin-related repositories and found more than 12 vulnerabilities, which it reported to the respective projects. The group is conducting AI-based security audits across key Bitcoin software, including wallets, cryptographic libraries and infrastructure, as part of building a so-called Bitcoin red team platform. It is also developing an open-source AI platform for Bitcoin software auditing.

A low-entropy vulnerability in the Coldcard hardware wallet has prompted bitcoin users to recheck how they generate seed phrases. Community focus has shifted toward physical entropy methods such as rolling dice rather than relying on device-based random number generators. After July 30, theft cases were publicly confirmed, with attackers brute-forcing private keys and stealing more than $100 million in BTC. Users are revisiting “don’t trust, verify” and exploring multiple offline methods to generate and cross-check BIP-39 seeds.