Bitcoin Core changed its security policy after a simple prompt enabled an artificial intelligence system to find bugs in its code. Jameson Lopp describes vulnerability discovery as having grown exponentially over the past two and a half months, with attackers already adopting uncensored AI models and researchers struggling to get much rest. He also breaks down the PHP remote code execution flaw, whose fallout was smaller than expected, and notes that 30% of Bitcoin addresses may have exposed keys. Lopp argues that security must now be handled in real time and built into the development pipeline rather than through annual audits, while a quantum-related upgrade could take years.
Source: Read the original article


