Skip to Content
Blockchain

We’re getting closer to being able to track stolen bitcoins

Cybersecurity experts have never been able to trace individual bitcoins, which is why it is so easy for cryptocurrency criminals to cover their tracks. A new algorithm could change that by revealing hidden patterns of Bitcoin money-laundering.

One of the big advantages of blockchain-based cryptocurrencies like Bitcoin is that the transactions are all recorded and publicly available. So it is always possible to see how much currency has been transferred from one account to another. (Although it’s not always possible to see who owns those accounts.)

But this transparency hides a dirty secret. While it is possible to see the flow of currency, bitcoins themselves are impossible to track.

That’s because bitcoins, and their smaller units satoshis, don’t exist as individual, identifiable items. They are not like dollar notes that have serial numbers. Instead, bitcoins are values that can be transferred from one address to another. The problem of tracking bitcoins is analogous to somebody depositing two $10 checks into a bank account, withdrawing $5 from a cash machine, and then asking which check the $5 came from. In the Bitcoin world, as in the real world, there is no way to answer that question.

And that causes problems when it comes to tracking the proceeds of crime. When bitcoins are stolen, the loot cannot be tracked and then reclaimed. Computer scientists have always held out hope that there may be a clever way to do so, but the algorithms developed so far have had limited success.

Enter Ross Anderson and colleagues at the University of Cambridge in the UK. These guys have built an algorithm adapting a 19th-century UK law that sets out a set of simple rules for dividing up money left over when a bank collapses. This law has become the basis for allocating money in a wide range of situations. And the researchers say that when they apply it to the public record of bitcoin transactions, it reveals remarkable patterns of criminal money-laundering activity that had been hidden until now.

The new algorithm is called Taintchain, and it has the potential to give law enforcement agencies an entirely new and powerful way to track the proceeds of cryptocurrency crime for the first time.

First some background. The theft of cryptocurrency is a big and growing business. In the first six months of 2018, some $761 million worth of cryptocurrency was stolen, according to the US cybersecurity firm CipherTrace. That’s over three times more than in the same period the year before.

The inability to track stolen funds efficiently is part of the attraction for cybercriminals. A common tactic, for example, is to place three stolen bitcoins in a wallet and add seven clean bitcoins. The 10 bitcoins are then split up and transferred to a large number of other accounts, and from there into still other accounts. Since there is no way to know which of the 10 bitcoins are tainted, the stolen currency quickly becomes diluted and lost. This process is called laundering.

One way to track this activity is to assume that all 10 bitcoins in the wallet are tainted and then follow the chain of transactions they are involved in. But that method ends up pointing the finger at an impractically large numbers of wallets, many of which will have unknowingly accepted the funds from other wallets.

Anderson and co have come up with a different tracking method based on the legislation known as Clayton’s Law. This established the first-in-first-out (FIFO) principle, which stipulates that when it comes to divvying up funds from an account, the first person to have paid in is the first person to be paid out. This principle has become enshrined in law all over the world as the fairest way to distribute funds when a bank or similar entity collapses.

The new Taintchain algorithm applies this principle to bitcoin wallets: if the first bitcoins paid into the wallet are stolen, then the first ones paid out are considered stolen too. So in the example above, where the first three bitcoins paid into the wallet were stolen, the algorithm assumes that the first three paid out are the stolen ones and then follows them to their next wallet, where it applies the same rule.

The Taintchain algorithm then displays the results in a way that allows suspicious patterns of behavior to appear.

This visualization process is difficult because of the sheer volume of transactions, but the team was able to identify a range of behaviors linked to money laundering.

For example, one pattern shows the way criminals divide the proceeds of a crime in a splitting pattern. “These may occur close to the time of a crime as criminals try to cover their tracks by feeding their loot into systems that divide their winnings into hundreds of tiny transactions,” say Ross and co.

This behavior is later followed by a collection pattern when the loot is regathered. “We observed similar patterns many times; in some of the instances, we were able to connect the collection address to illegal gambling sites,” say the team.

They also found other, more unusual patterns. One of them is a “peeling pattern” used by some exchanges and gambling sites. “Its operators would pool their money into a single wallet and then they would pay their customers successively, each time sending most of it to themselves at a change address,” say Ross and co.

Interestingly, in these cases the criminals tried to hide their identity by shuffling the transaction keys several times. But the algorithm is immune to this kind of sleight of hand because it looks only at the transfer of funds using the first-in-first-out process.

The work raises an interesting insight into the way criminals work when they launder money. The algorithm can only reveal these kinds of pattern if criminals behave in a way that mirrors the first-in-first-out principle.

Of course, that insight immediately suggests a way for malicious actors to hide their activity from Taintchain analysis by randomizing the way they pay out from wallets.

There is another factor that may become significant, which is the way the law applies to cryptocurrencies. The first-in-first-out principle generally applies to the distribution of money. But cryptocurrencies are not considered money in law.

That may change. If governments begin to recognize cryptocurrencies as money—and there is significant lobbying under way for that to happen—then a whole new set of financial laws will apply to cryptocurrency transactions.

One of these will be in the first-in-first-out principle. That will make the outcome of the Taintchain algorithm legally enforceable. (However, people who receive stolen bitcoins will not necessarily lose them, provided their transaction was made in good faith.)

That’s interesting work that has the potential to bring some law and order to the Wild West world of cryptocurrency transactions.

Ref: arxiv.org/abs/1901.01769  : Tendrils of Crime: Visualizing the Diffusion of Stolen Bitcoins

 

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.