Skip to content
- Bybit AI system stops massive scam withdrawals as hackers target crypto
- Exchange risk engine flags suspicious transactions before funds leave platform
- Automated monitoring shields thousands of users from potential crypto scam losses
A sharp rise in crypto-related cybercrime shaped the final months of 2025 as hackers intensified attacks on major exchanges. At the time, Bybit said that its artificial intelligence monitoring system had blocked or disrupted over 300 million dollars in suspected scam withdrawals. The number indicates the transactions that were intercepted prior to the outgoing of finances out of the platform.
There is internal information that indicated that the exchange had over 500 million withdrawal requests in the fourth quarter that were flagged. Over 4,000 users were immediately warned about taking risks or their transactions were frozen in real time. A significant number of account holders withdrew their transfers after reading the warnings. As a result, their money was not lost and their accounts were not in need of salvaging operation.
According to Bybit, the head of group risk control David Zong, the large portion of the 300 million was voluntary cancellations. This ensured that the exchange did not undergo clawbacks or reimbursement processes. Money was not taken out of the system and this minimized the operational pressure and possible customer conflict.
Meanwhile, the platform discovered 350 high-risk investment fraud addresses. This move has been protecting approximately 8,000 users against the risk of making withdrawal losses in the quarter. Moreover, Bybit said it had blocked over three million credential stuffing attacks during 2025. These robotic efforts are usually aimed at the login credentials in order to get unauthorized access.
In 2025, the industry had to pay off losses of crypto hacks in the amount of $3.4 billion. There was a tendency of hackers targeting large parties which are highly liquid. Thus, exchanges have increasingly been under pressure to reinforce preventative defenses instead of depending on investigations after an event.
Also Read: USAT Reserve Reveal Sparks Fresh Stablecoin Power Shift in US
Real Time AI Monitoring Limits Damage Before Funds Move
Bybit has developed its risk structure to study the behavior of withdrawal in real time. The system scans exchange information and marks some suspicious trends like mass withdrawals or suspicious destination wallets. When it identifies a high level of risk, it will either provide a warning signal or interrupt the transaction immediately. The degree of the threat identifies the course of action.
Furthermore, the framework enables the operations teams to blacklist high-risk addresses prior to the funds being received by them. The solution will change security to recovery reactive to security proactive. Consequently, the exchange will be trying to mitigate any losses as early as possible.
Industry Push for Preemptive Security Gains Momentum
Experts in the cybersecurity industry have encouraged cryptocurrency companies to use AI-driven surveillance. The significance of anomaly detection tools was previously mentioned by Deddy Lavid who is the co-founder and the CEO of the blockchain security firm, Cyvers. He added that real-time surveillance would assist in stopping hackers to penetrate platforms and withdraw finances.
The fourth quarter statistics of Bybit shows that exchanges have shifted their focus to early detection of threats. The automated alerts and address screening as well as the behavioral analysis represent a wider industry change. As attackers continue to optimize their attacks, preventive controls are becoming more and more dominant in terms of platform protection of user assets.
Also Read: Vitalik Buterin Sells 17,196 ETH in February as Ethereum Climbs
