The peer-to-peer (P2P) cryptocurrency platform NoOnes, used by over 2 million users, has announced a significant shift in its security model. The company has implemented an AI-powered system integrated directly into its escrow service to assess transaction risk before funds are transferred, moving away from traditional post-factum dispute resolution.
The new solution is not just an added filter but a fundamental change to the transaction process. The algorithm analyzes user behavior in real-time, considering factors such as atypical changes in trading activity, discrepancies in payment methods, and deviations in prices. If the system detects anomalies, it can suspend the transaction for additional verification or stop it entirely before funds are released from escrow.
According to NoOnes, a pilot program has already yielded strong results, reducing the proportion of disputed transactions by 28%. The system is reportedly able to identify up to 85% of potentially risky scenarios during the transaction execution phase. The company hopes this will reduce the workload on its support team and minimize conflicts between users.
This development comes amid a growing sophistication in cryptocurrency fraud. Analytics firms like Elliptic estimate that fraudulent transactions are becoming increasingly organized, using networks of accounts, automation, and social engineering. The U.S. Federal Bureau of Investigation (FBI) has also warned about the proliferation of crypto fraud schemes. Traditional trust models based on user ratings and transaction history are proving insufficient against these advanced threats.
While similar AI-driven behavioral scoring is commonplace in traditional fintech, adoption in the crypto P2P segment has been slower due to fragmented payment channels and varying user verification levels. NoOnes' move signals a broader industry trend toward proactive risk management. The platform believes such solutions could become the standard for P2P crypto trading, prioritizing security even if it means slightly slower transaction times for some users.
