Bybit says it helped recover 300 million dollars for thousands of users as crypto related fraud continued to surge across the industry. The exchange credited the results to its artificial intelligence powered fraud detection system, which is designed to step in before customers lose their funds.
Security Initiative Delivers Results
Sharing findings from its 2025 Security Initiative, Bybit stated that it intercepted 300 million dollars in impersonation scams and other fraudulent activity using a newly deployed AI based risk framework.
The announcement comes as digital asset crime remains a major concern. Data from Chainalysis shows that 17 billion dollars in crypto was lost to scams and fraud schemes in 2025.
According to the exchange’s report, 500 million dollars in withdrawals were flagged for review during the fourth quarter alone. Of that total, 300 million dollars was successfully intercepted and returned, safeguarding the funds of more than 4,000 users.
Over the same period, Bybit’s in house AI systems detected 350 high risk investment scam addresses using on chain analytics, protecting 8,000 individuals from potential withdrawal losses. The company also blocked more than 3 million credential stuffing attempts connected to account takeover schemes throughout 2025.
In addition, the platform automatically marked 350 suspicious wallet addresses and manually tagged another 600 through internal investigations. These actions prevented an estimated 1 million dollars in imminent fraud losses.
David Zong, Head of Group Risk Control at Bybit, said the firm aimed in 2025 to turn risk management into an active and intelligent safeguard by combining artificial intelligence with blockchain monitoring tools.
He explained that integrating AI driven on chain tracking with intelligence from partners such as TRM Labs, Elliptic, and Chainalysis not only protected Bybit customers but also helped identify patterns within fraudulent networks.
Three Tier Risk Protection Model
Bybit organizes potential scam threats into three escalating levels while maintaining smooth trading for legitimate users.
At the lowest risk level, the platform relies on large scale data analysis to spot unusual behavior, such as bulk withdrawals to newly created addresses. Automated surveys then assist the Risk Operations team in deciding whether to blacklist suspicious destinations.
For medium risk cases, real time alerts are triggered during withdrawal attempts. These may involve accounts flagged in credential stuffing databases or linked to questionable wallet addresses, prompting users to review transactions that could be influenced by social engineering.
At the highest risk level, wallet addresses tied to confirmed scams face immediate withdrawal suspension along with a mandatory one hour cooling off period.
The report concluded by outlining broader monitoring standards that could benefit the wider industry. These include an anti credential stuffing engine, real time AI pattern recognition to detect pig butchering schemes, an intelligence hub integrating tools from TRM Labs, Elliptic, and Chainalysis, and a cross chain tracing framework for tracking illicit funds from start to finish.#cryptonews https://t.me/coinsignalpublic https://coinsignals.net
