How the Wormhole hack worked: $320 million minted past a skipped check

In February 2022 an attacker created about 120,000 wrapped ether, worth more than $320 million, without putting a single real coin in. The flaw was an old verification routine that never checked who it was talking to.^[1] This case file covers how the Wormhole bridge was drained, who covered the loss, how part of it came back a year later, and the control that was missing.

What happened

How it worked

Before the bridge mints new tokens, it is supposed to confirm that a trusted group of validators has signed off on a real deposit on the other chain. That confirmation is the whole security of the system.

According to security analysts who reviewed the attack, Wormhole's Solana code relied on an older, deprecated function to perform part of that check. The function was meant to read information from a specific built-in system account, but it never confirmed that the account it was handed really was that system account. The attacker supplied a fake one, and the code accepted it as proof that the signatures had been verified.^[1]

From there, the bridge treated a forged message as an approved deposit and minted 120,000 wrapped ether. In everyday terms, a guard checked that a visitor was carrying an ID card but never looked at whose name was on it.

What it cost

At the time, it was one of the largest thefts in the history of decentralized finance.^[2][3] Ordinary users of the bridge did not lose their money, because Jump Crypto absorbed the loss and restored the reserve.^[3] The attacker was never publicly identified or charged.

A year later, part of the money came back. The attacker had parked much of the stolen ether in vaults on a lending platform called Oasis. On February 21, 2023, acting on an order from the High Court of England and Wales, Oasis used its ability to upgrade its own contracts to move the attacker's collateral to a wallet controlled by a party authorized by the court. Reports said Jump Crypto was behind the effort, and the net recovery was about $140 million.^[4]

The missing control

The missing control: retiring deprecated verification code before launch. The bridge's most important safety check depended on an outdated function that did not fully validate its inputs.^[1]

Outdated code is marked deprecated for a reason: the people who maintain it have found a better, safer way. A release checklist that required every security-critical check to use current, supported functions, and an independent review focused on how signatures are verified, would have flagged the weak routine. Once that check held, the attacker's fake account would have been rejected and nothing could have been minted.

What to do in your business

• Know what your key systems run on. Ask your developers or software vendors to list the libraries and functions behind logins, payments and approvals, and flag anything marked outdated.
• Make updates a release rule. Do not let new code or new features go live while they depend on components that their makers have told you to stop using.
• Get a second review on money-moving code. Anything that approves payments, withdrawals or refunds should be reviewed by someone who did not write it.
• Check the check. When a system says it verified something, such as a signature, invoice or identity, confirm it verified the right thing, not just that something was present.
• Plan for the bad day. Decide in advance who can pause a payment system, how customers will be told, and who covers losses, so you are not deciding under pressure.

Facts are drawn from court records, government reports, company statements and reputable reporting, listed below. People are named only where they were convicted, pleaded guilty or spoke publicly in an official role.