Dubai, May 12, 2026, 01:04 (GST)
- Binance reported that its AI tools blocked $10.53 billion in possible user losses, safeguarding over 5.4 million users between Q1 2025 and Q1 2026.
- Chainalysis puts losses from crypto scams at $17 billion in 2025, and says AI-powered schemes came in 4.5 times ahead of older scams for profit.
- KuCoin, a competing exchange, is cautioning its users about deepfake scams, “pig butchering” schemes, and address-poisoning wallet attacks. KuCoin
Binance says its AI-driven security tools stopped $10.53 billion in attempted crypto fraud over the past 15 months, underscoring how both exchanges and fraudsters are ramping up their tech. According to the company, it caught 22.9 million scam and phishing incidents just in the first quarter. Phishing—meaning bogus messages or sites aimed at stealing passwords or money—remains a major tactic.
It’s not just about how big the number is. Crypto fraud has evolved—moving beyond giveaway scams and spammy links, now leveraging deepfake videos, synthetic voices, bogus help desks, and bots capable of generating malware or baiting targets via chat.
Timing, then, is key. According to Chainalysis, impersonation scams soared more than 1,400% in 2025. Operations that had clear ties to AI vendors pulled in an average haul of $3.2 million each—substantially higher than the $719,000 for scams without those visible links.
By late 2025, Binance had rolled out upwards of 24 AI-driven projects and topped 100 models in production. The company says artificial intelligence is handling 57% of its fraud detection work, flagging fake payment evidence and catching questionable peer-to-peer messages. That’s translated to a 60%–70% drop in card-fraud rates versus industry standards.
Binance flagged its Binance Ai Pro setup, which puts AI-managed funds in a separate architecture, isolating them from user accounts and restricting withdrawals. The company says it’s also updating KYC identity screening to spot deepfakes and synthetic IDs.
Paul Ugbede Godwin, crypto analyst at EmageNewsDAO and MEXC, said in Tekedia that cybercriminals now “operate like multinational corporations,” targeting decentralized finance protocols, blockchain bridges, exchanges, and launching social-engineering schemes. His message is clear: AI has slashed attack costs and made it easier for more people to launch them. Tekedia
Competitors have also started issuing user alerts, though they haven’t disclosed blocked fund figures on par with others. KuCoin’s latest safety guide points to AI deepfakes, pig butchering scams—drawn-out cons that lure victims toward bogus investment sites—and address poisoning, a tactic where fraudsters slip a fake wallet address into a user’s transaction list.
The figures warrant some scrutiny. Reported directly by the company, they come without an independent audit or a full breakdown of how Binance arrived at its $10.5 billion headline, Intellectia pointed out. Binance, for its part, acknowledged that not every recovery can happen—the irreversible nature of blockchain transactions can make that impossible.
Binance faces a steeper credibility challenge than some rivals. The exchange and its founder, Changpeng Zhao, both admitted guilt as part of a U.S. criminal settlement in 2023. According to the Justice Department, Binance was used to channel illicit proceeds tied to ransomware, darknet operations, and online scams—all efforts to slip past authorities.
The industry’s picture isn’t summed up by one fraud-prevention figure. Exchanges keep ramping up their automated defenses. Scammers, though, are getting quicker, using their own automation to mimic real users and target people outside exchange platforms. Where this arms race lands—whether the next big win for crypto security happens inside trading tech, in customer education, or in those final moments before someone hits send—remains up in the air.
