Artificial intelligence has triggered a new wave of cyber threats. Binance says it has saved billions in user funds — with the help of AI.
As the use of artificial intelligence continues to grow, fraud attempts in the crypto sector are also increasing significantly. According to Binance, between the beginning of 2025 and March 2026, the company protected user assets worth 10.53 billion US dollars from fraud and scams. More than 24 AI-powered security initiatives and over 100 models were reportedly used for this purpose.
In total, the crypto exchange says it protected more than 5.4 million users from fraud cases and blocked around 36,000 suspicious wallet addresses. In the first quarter of 2026 alone, the company detected 22.9 million phishing and scam attempts.
AI Threats Are Increasing
“Fraud attempts and security vulnerabilities based on AI are increasing,” Binance said in a blog post. Deepfakes, fake platforms and phishing attacks in particular are being used in increasingly professional ways. According to the exchange, 57 percent of fraud detection is now handled by AI systems.
Binance says it has expanded its identity verification process to better detect AI-generated fake identities and deepfakes. The exchange is also making greater use of image recognition to identify forged payment confirmations. Real-time voice analysis is also intended to detect typical fraud patterns in chats and messages.
Conclusion
Crypto fraud is rapidly moving to a new level because of the development of AI. In the past, many scam schemes could be recognized by obvious mistakes and template-like messages. Now, attacks are becoming more personalized, visually convincing, and technically sophisticated.
Binance’s statement shows that major crypto exchanges are being forced to respond with the same technologies: automated analysis, image recognition, deepfake detection, and real-time monitoring of suspicious activity.
For users, the main takeaway is simple: even if a platform strengthens its protection, basic caution remains critical. Users should not trust links from messages, transfer funds at the request of “support,” accept payment screenshots as proof, or complete KYC outside official channels, concluded an expert from FORECK.INFO.
