Binance report says hackers used fake AI tools and phishing to steal $6.3 billion

Binance has published an initial report on the security failures behind the attack that led to the theft of $6.3 billion on 29 October. For high-risk PSPs, the useful part is not the size of the loss; it is the mechanics: compromised developer devices, fake job interviews, stolen credentials, and internal AI tooling used as part of the intrusion chain.

1. According to Binance, the attack started in June when a member of its wallet development team downloaded a tampered AI tool. The company said it detected the problem, replaced the device, and changed access credentials, but the attacker still obtained information about that employee’s role and about other members of the wallet team.
2. In September, a second software engineer searching for a new job downloaded a “malicious code repository” onto their computer after taking part in an interview they believed was legitimate. The process included a video meeting with the CEO of a fictitious company, which is a tidy reminder that social engineering does not need to be sophisticated when the target workflow is already messy.
3. On 20 October, the attacker used the compromised computer and an internal AI tool to take control of the system of one wallet developer and then the system of a second developer on the same team. On 29 October, the hacker used the second developer’s computer, together with stolen credentials from the first, to run a malicious script and launch the attack.
4. Binance said it is reimbursing all customers who lost funds in the attack. It also said it is increasing investment in AI-based detection, building a security research fund, and pursuing specific initiatives with leading AI companies.
5. The FBI attributed the attack to a North Korean hacker group, an accusation North Korea rejected. Binance’s report also included a direct acknowledgment from the company: “This should not have happened, and we apologize for the concern, the disruption and the uncertainty this caused to our users, partners and the industry as a whole.”