Binance says its artificial intelligence-powered security systems prevented $10.53 billion in user losses and blacklisted 36,000 malicious addresses between the first quarter of 2025 and the first quarter of 2026. The exchange said AI now drives 57% of its fraud controls.
The crypto platform disclosed the figures in a blog post on Monday, adding that it had protected more than 5.4 million users from fraud over the 15-month period. The exchange said it deployed over 24 AI-driven initiatives and more than 100 models during that span.
AI fraud controls scale as attack sophistication rises
The barrier to entry for running scams has fallen. What once required technical knowledge can now be executed cheaply and at scale, according to Binance.
Threat actors are using AI to accelerate phishing campaigns, create deepfakes, clone voices, and impersonate platforms across chat applications. Social engineering has been amplified to a level not seen before, the exchange said.
Scams and exploits have been a persistent problem in crypto. The FBI reported in April that US citizens lost $11 billion worth of crypto to scams, with impersonation of government officials or crypto companies being a common tactic.
| Metric | Q1 2025 – Q1 2026 | Q1 2026 Alone |
|---|---|---|
| User funds protected | $10.53bn | $1.98bn |
| Users protected | 5.4m | Not disclosed |
| Malicious addresses blacklisted | 36,000 | Not disclosed |
| Scam attempts intercepted | Not disclosed | 22.9m |
Computer vision and real-time language analysis deployed
Binance said it has implemented computer vision to detect fake payment proofs and real-time language analysis to spot scam patterns. The exchange also integrated AI into identity verification to counter deepfakes and synthetic identities.
AI-driven decisioning now powers 57% of fraud controls at Binance. The exchange said this contributed to a 60% to 70% reduction in card fraud rates compared to industry benchmarks.
In the first quarter of 2026 alone, the exchange intercepted 22.9 million scam and phishing attempts, saving $1.98 billion in user funds.
The crypto fraud arms race accelerates
The shift to AI-based attacks has been swift. Highly organised threat actors have adopted the technology to create more sophisticated exploits, whilst the cost of launching scams has dropped.
Binance’s figures suggest the defence side is scaling as well. Over 100 models deployed across 24 initiatives in 15 months is a significant engineering commitment.
The question is whether the fraud controls scale faster than the fraud. The exchange’s numbers show billions protected, but they also show tens of millions of attempts per quarter. That volume is not slowing.
This article is for information purposes only and does not constitute investment advice. Readers should not act on any information contained here without first consulting an authorised financial adviser. Past performance is not a reliable indicator of future results.
