Key Takeaways
- Binance reported that its AI-powered security systems helped prevent $10.53 billion in potential user losses from early 2025 through Q1 2026.
- Google identified a zero-day exploit that it believes was developed with AI.
- Chainalysis estimated that crypto scams and fraud stole $17 billion in 2025.
- AI now runs through exchange defense, fraud automation and exploit development.
Crypto’s AI arms race is no longer theoretical.
On one side, exchanges are using machine-learning systems to scan wallets, flag suspicious behavior and stop scams before funds move.
On the other hand, cybercriminals are increasingly using the same technology to build smarter phishing attacks, automate fraud and develop more advanced exploits.
Binance just offered one of the clearest signs yet of how large that battle has become.
Crypto’s Strongest Defense and Fastest-Growing Threat
The exchange said its AI-powered security systems helped prevent $10.53 billion in potential user losses from the start of 2025 through the first quarter of 2026, covering more than 5.4 million users globally.
The disclosure arrived almost simultaneously with a warning from Google, which said it identified a zero-day exploit that its researchers believe was developed with the help of AI.
Together, the reports paint a picture of an industry entering a new phase — one in which artificial intelligence is rapidly becoming both crypto’s strongest defense and one of its fastest-growing threats.
Binance Says AI Prevented Billions in Potential Losses
According to Binance, the company’s fraud-prevention infrastructure now relies on more than 24 AI-driven security initiatives and over 100 machine-learning models operating across scam detection, transaction monitoring, address screening and risk analysis.
The exchange said those systems intercepted 22.9 million scam and phishing attempts during Q1 2026 alone while helping safeguard roughly $1.98 billion in user funds over the same period.
Binance also reported:
- More than 36,000 malicious wallet addresses have been blacklisted.
- Over 9,600 real-time warnings are issued daily.
- AI is now powering more than half of its fraud-detection stack.
The numbers come from Binance and have not been independently verified.
Still, they offer a glimpse into how heavily major exchanges now rely on automation to manage increasingly industrialized fraud.
Crypto Scams Are Becoming More Sophisticated — and More Automated
Binance’s numbers arrive amid a broader surge in automated crypto fraud.
Blockchain analytics firm Chainalysis estimated that scams and fraud stole approximately $17 billion in 2025.
The company also reported that AI-assisted scams generated significantly more revenue than traditional scam operations.
According to Chainalysis, scam networks are increasingly using:
- AI-generated phishing messages.
- Deepfake impersonation videos.
- Automated scam infrastructure.
- Phishing-as-a-service kits.
- Large-scale laundering operations.
The economy is changing quickly.
Generative AI tools make it easier for attackers to localize scams in multiple languages, imitate real support agents, produce convincing fake websites and test attack methods on a much larger scale.
Binance Research separately found that intercepted phishing and scam attempts rose 54% quarter-over-quarter and 209% year-over-year during Q1 2026.
In other words, exchanges are facing not just larger fraud losses but also dramatically higher overall attack volume.
Google’s AI Warning Shows the Threat Is Expanding Beyond Crypto
The story became even more serious after findings released by Google Threat Intelligence Group.
Google Threat Intelligence Group said it found a threat actor using a zero-day exploit that it assesses was developed with AI.
According to the report, the exploit targeted a widely used open-source web administration tool and was designed to bypass two-factor authentication protections.
The exploit reportedly appeared inside a Python-based attack script intended for broader mass exploitation.
Google said early detection may have disrupted the actor’s plans before the exploit could spread widely.
But the warning itself mattered because it signaled something cybersecurity experts have increasingly feared: AI is moving beyond simple phishing automation and into exploit development itself.
The company also warned that attackers are increasingly experimenting with:
- AI-assisted vulnerability discovery.
- Polymorphic malware.
- Autonomous attack behavior.
- Defense evasion systems.
- Supply-chain attacks involving AI infrastructure.
Google added that state-linked threat actors connected to China, North Korea and Russia have shown growing interest in AI-assisted cyber operations.
Crypto Is Especially Vulnerable to AI-Driven Fraud
Crypto is unusually exposed to automated fraud because a successful attack can settle before any human review catches up.
Phishing campaigns, fake support messages, wallet-drainer sites, poisoned addresses and malicious approvals already move quickly.
AI gives scam operators cleaner messages, faster localization, cheaper code testing and more flexible infrastructure rotation.
For exchanges, the defensive answer is also automation.
Address blacklists, behavioral signals, transaction-risk scoring and real-time alerts are becoming part of the platform itself.
That creates a stronger safety perimeter around exchange accounts.
It also shifts more judgment into private risk models.
A blacklist can block a known scam address.
A risk engine can delay, flag or restrict activity through criteria users may never see.
As AI becomes embedded in exchange security, the question will move beyond detection rates.
Platforms will have to show how their systems classify risk, how false positives are handled and how much control automated models have over user activity.
Security Is Becoming a Competitive Advantage for Exchanges
For years, exchanges competed primarily on fees, token listings and liquidity.
A platform claiming it prevented billions in fraud losses is also making a broader pitch to users and institutions: that it offers safer infrastructure in an increasingly hostile digital environment.
That matters particularly as Wall Street firms, institutional investors and regulators push deeper into crypto markets.
But Google’s warning also reinforces a difficult reality for the industry.
The same AI systems helping exchanges stop attacks are also helping criminals scale them.
Which means the next phase of crypto security may not be about building stronger defenses alone.
It may become a constant race between competing AI systems operating on both sides of the market.
Recommended Secure Partners
PhD, researcher and writer exploring AI, blockchain, and the philosophy of tech, with a focus on DeScAI, governance, and trust.
