Bybit Intercepts $300 Million in Fraudulent Withdrawals through AI Security Framework in 2025
Bybit announced on February 27 that it intercepted $300 million linked to fraud in 2025. The exchange's AI-driven risk control framework, including a Three-Tier Withdrawal Anti-Fraud Defense System, identified suspicious activities and protected over 4,000 users. In total, $500 million in withdrawals were flagged, with significant measures against scams including blocking over 3 million account takeover attempts and collaborating with external intelligence providers to enhance security.
Theia Market Signal Identification - AI Assisted
Bybit disclosed that it intercepted $300 million in fraudulent activity during 2025, utilizing a new AI-powered risk control framework. The Three-Tier Withdrawal Anti-Fraud Defense System categorizes risks into three levels, enabling interventions based on detected anomalies.
Of the $500 million in flagged withdrawals, $300 million was recovered, protecting over 4,000 users. Bybit also identified 350 high-risk addresses and blocked more than 3 million credential stuffing attempts. The measures included collaborations with TRM, Elliptic, and Chainalysis to enhance monitoring and security across the platform.
Comments