Poly Network Hack: How $610 Million in Crypto Was Returned

Originally published 2021-08-16.

On 10 August 2021, an attacker drained roughly $610 million in cryptocurrency from Poly Network, a platform that lets users move tokens between different blockchains. At the time it was the largest theft in the history of decentralised finance. Then the story took a strange turn: within a day the attacker began sending the money back, Poly Network nicknamed them "Mr. White Hat", offered a $500,000 bug bounty and even a job as chief security advisor, and by late August the company said all user assets had been recovered. Here is how the hack worked, how the return unfolded, and what it says about security in crypto.

What Poly Network is and what was taken

Poly Network is a cross-chain bridge. Blockchains such as Ethereum, BNB Smart Chain (then Binance Smart Chain) and Polygon cannot talk to each other directly, so bridges use smart contracts to lock tokens on one chain and release equivalent tokens on another. To do that, the bridge's contracts hold large pools of users' funds, which makes them an attractive target.

The attacker moved funds out of Poly Network contracts on all three chains into wallets they controlled. Reported totals settled at a little over $610 million, spread across many tokens including Ether, Binance Coin, stablecoins and wrapped versions of Bitcoin. That beat the previous record for a crypto theft, the $534.8 million taken from the Japanese exchange Coincheck in 2018.

How the attack worked

Bridges rely on special accounts, sometimes called keepers, that are trusted to sign off on cross-chain transfers. According to the post-incident analysis reported by Infosecurity Magazine, the attacker exploited a flaw in how Poly Network's contracts handled cross-chain messages to change the keeper role on two contracts. Once they had replaced the keeper with their own key, they could authorise withdrawals of any size.

Crucially, the underlying blockchains were not "hacked". Ethereum and the others did exactly what they were told. The weakness lay in the code Poly Network had written on top of them. Tom Robinson, co-founder of the blockchain analytics firm Elliptic, told the BBC that "whenever a human being writes code, there's a chance they will make a mistake." Smart contracts can hold enormous sums, are public for anyone to study, and are hard to change once deployed, so a single bug can be catastrophic.

From theft to "Mr. White Hat": the timeline

• 10 August: Funds are drained from Poly Network contracts on Ethereum, BNB Smart Chain and Polygon. Poly Network publicly appeals to the attacker and to exchanges and miners to blacklist the addresses.
• 11 August: The stablecoin issuer Tether freezes about $33 million in USDT held by the attacker. The attacker, posting messages embedded in blockchain transactions, says the hack was meant to expose the flaw, and that they always planned to return the money. Returns begin.
• 13 August: Hundreds of millions of dollars have come back, and most of the rest is moved into a multi-signature wallet that needs keys from both the attacker and Poly Network to unlock.
• Mid-August: Poly Network calls the attacker "Mr. White Hat", offers a $500,000 bounty and invites them to become its chief security advisor.
• 23 to 25 August: The attacker hands over the final private key. Poly Network says all user assets moved out during the incident have been fully recovered.

The $600 million Poly Network hacker has published part one of a "Q&A":#polynetworkhack pic.twitter.com/3y1JQnHe50

— Tom Robinson (@tomrobin) August 11, 2021

As the tweet above shows, the attacker even published a question-and-answer style explanation of their motives, written into blockchain transactions so anyone could read it. One line from it asked whether people should not "learn something from those hacks".

Was it really a white-hat hack?

In security, a white-hat hacker finds flaws with permission, or reports them responsibly so they can be fixed before anyone is harmed. A black-hat hacker exploits flaws for personal gain. Poly Network's decision to call the attacker a white hat drew criticism from many security experts, because the attacker did not ask permission or report the bug quietly; they took the money first.

Several practical pressures may explain why the funds came back:

• Everything was visible. Public blockchains record every transaction. Analytics firms and the wider crypto community were tracking the stolen funds in real time.
• Cashing out was hard. Turning hundreds of millions in traceable tokens into ordinary money without passing through exchanges that check identities is extremely difficult.
• Some funds were frozen. Tether's freeze showed that parts of the loot could be blocked outright.
• Investigators reportedly had leads. Security firms said they had identified some of the attacker's digital traces, though no identity was confirmed publicly.

Robinson suggested to the BBC that the attacker may have realised how much attention they were getting and that they would be watched wherever the funds moved. Whether the motive was principle, fear of capture, or both remains unknown. No arrest was made public.

Can a company forgive a theft?

Offering a reward does not make an unauthorised attack legal. In most countries, prosecution for computer fraud or theft is a decision for law enforcement and prosecutors, not the victim company. Poly Network said it would not pursue the matter, but that promise did not bind any government.

What happened next, and the bigger picture

After the incident, Poly Network launched a public bug bounty programme on the Immunefi platform, offering up to $100,000 for critical vulnerabilities. That is the normal, legal route for white-hat researchers: find the flaw, report it privately, and get paid without touching anyone's funds.

The episode turned out to be an early warning. Cross-chain bridges went on to suffer several of the largest crypto thefts ever recorded in the following years, and in most of those cases the money was not returned. Poly Network's story is unusual mainly because of its ending.

Lessons for anyone using DeFi

• Code risk is real. Funds held in smart contracts are only as safe as the code and the people who control its keys. Audits help but do not guarantee safety.
• Bridges concentrate risk. Because they hold large pools of tokens, bridges are frequent targets. Keeping large amounts in transit or in bridge contracts adds exposure.
• Recovery is not normal. The Poly Network refund was a rare outcome. Most stolen crypto is never recovered, so do not count on a happy ending.
• Check the project's security history. Look for published audits, a bug bounty programme and clear communication about past incidents before trusting a platform with money.

The Poly Network hack is remembered less for its size than for its bizarre resolution. It showed both the weakness of complex financial code and the strange transparency of blockchains, where a thief can steal in public, explain themselves in public, and end up giving it all back.

Sources: Wikipedia: Poly Network exploit; Infosecurity Magazine: Poly Network hacker returns funds; original report via Futurism; tweet by Tom Robinson (@tomrobin).