TRM Labs: KYC for AI Agents Must Start With the Human Behind Them

North Korean hackers stole roughly $6.3 billion from Binance last week by duping employees through a fake job interview and exploiting an internal AI tool, with most of the funds vanishing within 20 minutes. The breach has intensified scrutiny of how AI agents that initiate payments should be verified, and TRM Labs argues that know-your-customer checks for such agents must begin with the human behind them. The incident matters because it shows how rapidly criminals can move billions once they are inside a platform, a risk that grows as autonomous software takes on more payment duties. What to watch next is how exchanges and financial institutions adapt their KYC frameworks to cover AI agents while keeping accountability tied to a real person.