**[Bybit Successfully Intercepts $300 Million in Fraudulent Funds in 2025]** BlockBeats News, February 27 — Bybit announced the results of its 2025 security plan. According to the official statement, the platform successfully intercepted approximately $300 million in funds related to scams, impersonation, and fraudulent activities in 2025 through its new AI-driven risk control framework. Bybit stated that its "Three-Tier Withdrawal Anti-Fraud Defense System" is based on a dynamic risk grading mechanism: - **Tier 1 (Low Risk):** Identifies abnormal patterns (e.g., concentrated transfers to new addresses) through big data heuristic models and automatically issues questionnaires to blacklist high-risk addresses in advance. - **Tier 2 (Medium Risk):** If an account is linked to leaked databases or suspicious addresses, real-time alerts are triggered during withdrawals, prompting users to review transactions and prevent social engineering scams. - **Tier 3 (High Risk):** For confirmed fraudulent addresses, real-time withdrawal interception is implemented, along with a mandatory one-hour cooling-off period. **Key Data (Q4 2025)** - Out of $500 million in suspicious withdrawals, $300 million was successfully intercepted and recovered, involving over 4,000 users. - AI algorithms identified 350 high-risk investment scam addresses, protecting 8,000 users. - Over 3 million credential stuffing attacks were intercepted throughout the year. - 350 addresses were automatically flagged, and 600 addresses were manually flagged, preventing approximately $1 million in potential losses.