Crypto fraud reached new heights in 2025, with illicit transactions crossing $154 billion globally, a 162% jump from the year before. Artificial intelligence has emerged as the industry’s primary line of defence, powering real-time transaction monitoring, phishing detection, identity verification and anti-money laundering checks across major platforms. As illicit crypto transactions surged 162% year-on-year, exchanges are turning to artificial intelligence to catch fraud before it happens. Alan Xin, Head of Blockchain Risk Control at Bybit, explains how.

Q2. What kinds of fraud are crypto users most exposed to today?

Users face phishing attacks, fake investment schemes, identity theft, account takeovers, and social engineering, and the tactics keep evolving. What makes this especially serious is that blockchain transactions are usually non-reversible once completed. That changes the entire approach to security. Preventing fraud before it happens matters far more than trying to recover funds afterwards.

Q3. How is AI changing the way exchanges detect fraud?

Real-time data analysis is one of the most effective tools we have. Exchanges process millions of transactions daily across multiple   markets and blockchain networks, a volume that simply cannot be monitored manually. AI allows platforms to analyse this data within fractions of a second, scanning transaction patterns, user behaviour, device information and account activity to flag unusual activity that may indicate illegal behaviour.

Q4. Can you give an example of what that monitoring looks like in practice?

By continuously tracking transaction patterns and login activity, our systems can quickly identify actions that deviate from a user’s normal behaviour, such as an unusually large withdrawal, for instance, or a login attempt from an unfamiliar location. Depending on the risk level detected, the system may pause the transaction or request additional verification before it proceeds. Catching these warning signs early significantly reduces the risk of financial loss for users.

Q5. Phishing remains one of the most common scams. What role does AI play there?

Cybercriminals constantly create fake websites, fraudulent emails, and deceptive messages to trick users into revealing sensitive information. AI helps counter this by identifying suspicious links, recognising scam patterns, warning users about potential risks, and continuously learning to detect new tactics as attackers adapt. It’s an ongoing arms race, and AI helps us keep pace.

Q6. How is AI being applied to identity verification and KYC?

Know Your Customer processes remain essential for regulatory compliance and preventing financial crime, and AI-powered verification is making these processes far more efficient. By analysing identity documents with greater speed and accuracy than manual methods, AI strengthens security standards while making verification faster and smoother for legitimate users.

That builds greater confidence for institutions looking to participate in the digital asset ecosystem.

Q7. Money laundering is another major concern in crypto. How does AI help address it?

Criminals often move illicit funds across multiple wallets and exchanges to disguise their origins. AI can analyse blockchain data, track suspicious fund movements, and identify patterns that may indicate money laundering. That allows exchanges and compliance teams to detect suspicious transactions faster and strengthen their overall compliance efforts.

Q8. Looking ahead, what will determine whether crypto continues to grow sustainably?

Growth doesn’t depend on technology alone; it depends on trust. Users need confidence that

their funds, personal information, and transactions are secure. By adopting AI-powered fraud detection, platforms can reduce risk, strengthen security standards and improve investor confidence. As the industry matures, AI will remain central to that effort and to supporting long-term, sustainable growth.