The podcast episode centers on cybersecurity vulnerabilities and recent hacks, particularly the Drift Protocol exploit and the Claude Code source code leak. The discussion begins with the Drift Protocol hack, estimating losses of around $250 million, and explores potential attribution to DPRK-linked actors, highlighting the increasing sophistication of their methods, such as supply chain attacks. The hosts then shift to the Claude Code leak, examining the implications of AI agents checking in code and the rapid pace of code evolution. They note the surprising value of leaked code in the age of AI, and discuss the potential for open-source models to benefit from the leaked tricks and harnesses. The hosts also share personal anecdotes related to phishing attacks and business email compromise, and discuss strategies for mitigating security risks.
Sign in to continue reading, translating and more.
Continue
