Binance's latest blog post revealed that over the past year, the platform has used more than 100 AI models to intercept over $10 billion in fraudulent activity. The company stated that from Q1 2025 to Q1 2026, these measures have covered and protected over 5.4 million users.

22.9 million attacks intercepted in the first quarter of 2026.

The company disclosed that in the first quarter of 2026 alone, the platform blocked 22.9 million fraud and phishing attempts and helped protect approximately $1.98 billion in user funds. Binance stated that its current protection system relies primarily on large-scale AI deployment, real-time monitoring, and account security design.

Malicious addresses and real-time alerts are being increased simultaneously.

Binance stated that it has blacklisted over 36,000 malicious addresses and issues over 9,600 real-time alerts to users daily to identify evolving scam tactics. In addition to system interception, the platform is also promoting user education.

The company stated that its security education programs regarding account takeover risks reached over 179,000 users in the first quarter of 2026. These programs primarily serve to alert users to unusual logins, phishing links, and social engineering attacks.

Exchanges use AI as a risk control tool

Binance stated in a blog post that as threat tactics become more automated, the platform is using AI to identify anomalous behavior, enhance real-time response, and as part of user fund protection. This statement comes from a unilateral disclosure by the company, and the article did not further clarify the specific statistical criteria for "intercepting over $10 billion in fraud."

Additional information:The original article also mentioned that Binance recently expanded its commodity derivatives contracts to cover assets such as crude oil, natural gas, gold, and silver. In January of this year, the platform launched perpetual contracts for gold and silver settled in USDT.