
S0:E66: Binance's AI Agent Got Hijacked — Plus North Korea's Fake Job Interview Malware w/ Richy
A gang reportedly used a fake North Korean-style job interview to plant malicious software on two Binance engineers' computers — and walked away with enough that the exchange had to reimburse 90% of the stolen wallets out of its own reserve fund. Jess is out this week (and for a few more), so Richy's back in the chair with Jake to dig into the New York Times' Binance postmortem: how threat attribution actually works when everyone's pointing at North Korea, why Binance's own SAFU fund raises more questions than it answers, and whether an AI tool was the privilege-escalation step that got attackers onto a developer's machine.
Then: North Korea's classic fake-job-interview malware playbook, why 'if it sounds too good to be true, it is' still works on smart people, and why almost nobody is seriously testing their AI agents for prompt injection yet.
Did we make you smarter? Tell a friend. Did we make you mad? Tell your CISO.
Sources: Binance hack postmortem (New York Times): https://www.nytimes.com/2026/11/03/technology/binance-hack-report.html
join our discord: https://discord.gg/wYJnTpczV
More episodes from "Breach Please"



Don't miss an episode of “Breach Please” and subscribe to it in the GetPodcast app.








