Crypto security is turning into an AI arms race as agents may overwhelm compliance teams
AI agents and automated payments could reach a scale that crypto monitoring systems built for human-paced markets cannot handle, Elliptic CEO Simone Maini warned.
What to know:
— Crypto security’s biggest emerging risk is not larger hacks, but AI-driven financial activity occurring at a speed and scale humans cannot handle, Elliptic CEO Simone Maini said.
— As AI makes it cheaper and easier for bad actors to run large-scale hacks, scams and fraud, firms like Elliptic are responding with their own AI agents to analyze blockchain data in real time, turning crypto security into an automated arms race, she said.
— Elliptic raised $120 million from investors including Nasdaq and Deutsche Bank to build out its agentic compliance system as banks and institutions are pushing deeper into digital assets.
Crypto’s next security problem may not be bigger hacks, but financial activity driven by artificial intelligence (AI) happening at a scale compliance teams can’t keep up with.
That shift is forcing crypto security firms to rethink how compliance works in an increasingly automated financial system, said Simone Maini, CEO of blockchain analytics firm Elliptic.
«When you think about agentic commerce, we’re thinking about the sheer volume of transactions and events that need to be monitored as growing exponentially,» she told CoinDesk in an interview.
The problem, she said, is that today’s compliance systems still rely heavily on manual review. Analysts investigate alerts, trace wallets and flag suspicious activity. That model starts breaking down once financial activity starts happening continuously and at machine speed.
«There simply aren’t enough compliance analysts specializing in digital assets in the world to be able to keep up with these volumes,» Maini said.
This matters for crypto because stablecoins, tokenized assets and AI-driven payments are increasingly seen as the next growth areas for the industry, while traditional finance push deeper into digital assets. As more banks and asset managers move onchain, compliance and monitoring systems will also need to scale alongside them.
The AI arms race in crypto security
Earlier this week, Elliptic raised $120 million backed by investors including Nasdaq and Deutsche Bank to build what Maini described as «agentic» compliance system: AI-driven tools designed to automate transaction monitoring and investigations that today overwhelm compliance teams.
«As transaction volume is growing, cost per alert handling, cost per investigation is falling,» she said. «For us, what we’re essentially doing for our customers is inverting that cost curve in compliance.»
At the same time, the same AI tools helping compliance teams are also making attacks cheaper and easier to scale.
«AI can equip the bad actors with the ability to perform hacks, scams at a scale that they couldn’t do when they were reliant on writers,» Maini said.
That dynamic is turning crypto security into an increasingly automated arms race. As attackers use AI to lower the cost of phishing, exploits and fraud campaigns, security firms are deploying their own models to identify illicit activity faster and across larger datasets.
For Elliptic, that includes using AI agents internally to collect blockchain intelligence, attribute wallets and detect suspicious patterns in real time.
Despite a recent wave of high-profile hacks, Maini said the industry has made more progress than many outsiders realize.
She argued that monitoring tools have matured significantly over the past decade, especially given the growth in transaction volume and the expansion into stablecoins, decentralized finance and tokenized assets.
«The Pandora’s box is open when it comes to the institutionalization of digital assets,» she said. «That much is clear.»