Company News

How Chain, Our AI Workspace, Helps Investigators Fight Crime

The biggest hack in crypto history is also likely the most complex. Binance’s attackers are weaponizing automation in their race to launder billions of dollars in stolen crypto.

But Chainalysis’ first responders have our newest weapon: Chain. With our AI-native workspace, we’re tracing stolen funds as they spread over tens of thousands of addresses. The launderers want Binance’s money to outpace the tracers. With Chain, we’re not letting them.

Chain is the culmination of everything we’ve built over 12 years of following the money. It brings together Chainalysis attributions, the analytical expertise behind millions of investigations, and the flexibility of conversation, blending it together wherever our investigators and researchers work.

So when hackers began draining billions from the world’s largest crypto exchange around 4:00pm on a Thursday, our team, armed with Chain, set off.

Chain has three core capabilities useful in a crisis, as well as in everyday work: it will converse with you, it will create with you, and it will codify your labor-saving workflows. Our team used all three in responding to Binance.

Converse

Chain puts Chainalysis intelligence to work through conversation. Ask it a question, and it answers using the breadth of our on-chain knowledge.

Our investigators held round-the-clock conversations with Chain to understand how the Binance hackers were laundering funds. Their first task was determining where to look: the attackers initially held the stolen crypto across nearly a dozen different blockchains. And they were making this map even harder to navigate by moving the funds over bridges.

Rather than manually search through different blockchains, our investigators simply asked Chain: find corresponding bridge transactions.

“The data was really strong,” one of our investigators said. “We didn’t need external APIs, we didn’t need explorers, we didn’t need custom scripts. We could just say, ‘Hey Chain, go find the corresponding bridge transactions,’ and it would go and do it.”

Create

Chain turns intelligence into the outputs investigators need. Describe what you want—a graph, a report, a brief—and Chain helps build it.

Our research team utilized Chain’s creative output to chart out the impact of the Binance hack. They wanted to compare this singularly large hack to the long history of North Korean-linked thefts.

Because Chain is stocked with over 12 years-worth of attribution data, it readily puts historical data points like these into context. That’s exactly what it did here. The research team came with a request—what share of annual losses to crypto hacks are North Korean hackers responsible for?— and worked collaboratively with Chain to plot out an answer.

“When time is of the essence, you need to be able to collate facts quickly, correctly, and from a wide variety of sources, ” said Head of Research Eric Jardine. “Chain let us take our corpus of existing DPRK-linked hacks and flesh out other events from publicly reachable online sources. This rapid, iterative process allowed us to create a definitive dataset of DPRK-linked events, from which we could understand the trend in events and the sheer enormity of North Korea’s role in the hacking ecosystem.”

Codify

Tracing frenetic flows one wallet at a time can’t keep up when attackers use automation to move billions of dollars. Chain helps investigators turn their own methods into Automations that handle repeatable tracing work.

Our investigators noticed the hackers’ laundering practices fit recognizable patterns. They were breaking huge sums into manageable chunks. Then they were moving those chunks through known laundering routes, like DeFi protocols that swap one token for another.

Those patterns are where Chain made the difference. Once investigators understood a pattern, they could build an Automation in Chain to handle the repetitive tracing work. They applied that workflow to one deliberately convoluted laundering flow in which the attackers were swapping stablecoins back and forth.

“We built an automation in Chain that said: take the address, find the swap to one stablecoin, find the recipient of that stablecoin, find the swap back to another stablecoin, and then figure out where it went,” one investigator explained. He called this automated workflow one item of Chain’s “recipe book” for tracing the flow of funds.

A reactor graph of on-chain flows that Chain can codify into a workflow.

Chain’s recipe book of automations give us and our customers — among them exchanges and law enforcement — the best chance to disrupt threat actors. Our private sector partners can better manage their risk exposure on their platforms, and law enforcement can develop intelligence, secure evidence and know where to seize assets.

What’s next for Chain

Our investigators now use Chain in their day-to-day work. That work is strengthened by partnerships like the Digital Crime Task Force, which we launched with TRON and Tether to proactively identify and disrupt financial crime.

The capabilities our investigators relied on during the Binance response are what we’re now excited to bring to customers. With Chain, compliance, risk and investigations teams can get answers they can act on, build what they need, and turn their best methods into Automations that scale across the organization.

As bad actors adopt new technology, the teams working to stop them need to move just as fast. Chain is now in limited preview. Register for early access to learn more.