TheBlock
Binance says its AI-driven security systems have blocked more than $10.5 billion in cumulative user losses from scams and phishing since 2023, including $1.98 billion safeguarded in Q1 2026 alone.
In the first quarter, the exchange flagged 22.9 million scam and phishing attempts targeting its users, ranging from credential-harvesting pages to address-poisoning schemes and pig-butchering operations. The $1.98 billion figure is Binance's own estimate of the funds that would have been lost absent intervention.
Why it matters
The numbers are the largest self-reported scam-prevention dataset any major exchange has published, and they arrive as AI-generated fraud tooling lowers the technical bar for would-be scammers. Binance Research has separately estimated that AI is currently 2x better at exploitation than detection, and that AI-enabled scams are 4.5x more profitable than traditional variants — a gap that platforms are now forced to close with the same class of tooling.
Market impact
For users, the read is operational rather than directional: exchange-side AI defenses have become a primary line of defense against a threat environment where the cost to attack has fallen faster than the cost to defend. Watch the next quarterly disclosure for whether the volume of attempts continues to scale or whether detection rates begin to close the gap with the exploitation side of the AI ledger.
Frequently asked questions
-
How much in user losses did Binance say it prevented?
Binance said its AI security systems have blocked more than $10.5 billion in cumulative user losses from scams and phishing since 2023, including $1.98 billion safeguarded in Q1 2026 alone.
-
How many scam attempts did Binance flag in Q1 2026?
The exchange said it flagged 22.9 million scam and phishing attempts in Q1 2026, including credential-harvesting pages, address-poisoning schemes, and pig-butchering operations.
-
Is the $10.5 billion figure independently verified?
No. The figure is Binance's own estimate of funds that would have been lost absent intervention. It is self-reported and has not been audited by an external party.
-
What did Binance Research say about AI and scams?
Binance Research has estimated that AI is currently 2x better at exploitation than detection, and that AI-enabled scams are 4.5x more profitable than traditional variants.
-
What types of scams were targeted in Q1 2026?
Binance cited credential-harvesting phishing pages, address-poisoning schemes, and pig-butchering operations among the categories flagged in Q1 2026.
CoinDesk
Crypto Capital Venture
CryptoSlate
Lookonchain
CoinTelegraph
WuBlockchain
WatcherGuru