Smashing Security podcast #475: JadePuffer – the AI that ran a ransomware attack all by itself

This episode covers three security issues: Sysdig's analysis of "Jade Puffer," a fully autonomous AI-driven ransomware that gained access via an unpatched internet-facing flaw, stole credentials and API keys, encrypted a production database but failed to deliver a usable ransom due to implementation errors; multiple cases in Japan where teenagers used AI chatbots to build attack tools — including one that cancelled 46,812 streaming accounts and another that exfiltrated 7.25 million records — illustrating how AI lowers the skills barrier for cybercrime; and a disclosed weakness in Apple's Hide My Email feature that can link masked addresses to original inboxes, posing risks for privacy and vulnerable users.