Chainalysis estimates crypto scams caused about $17 billion in losses in 2025 worldwide, driven by AI tools and impersonation tactics, according to the report (Ed. note: this is a record high).
Average scam payments rose to roughly $2,800 in 2025 from about $800 a year earlier, a 253% increase. “On a time-weighted basis, you get faster scale and better believability,” said Eric Jardine in the report.
Scams linked on-chain to AI vendors averaged about $3.2 million per operation, roughly 4.5 times larger than non-AI operations. Scammers used face‑swap tools, deepfakes, and large language models often sold via Telegram channels.
Government impersonation attacks using deepfaked images grew more than 1,400% year over year. One mass phishing campaign impersonated toll alerts and sent as many as 330,000 texts in a single day, tied to groups identified as Darcula or the Smishing Triad, with phishing kits costing under $500.
Long-term “pig butchering” scams still yield higher individual losses by building trust over months. A San Jose woman later used ChatGPT to confirm a romantic partner was a scammer after losing nearly $1 million in cryptocurrency.
Impersonators increasingly move funds through DEXs, DeFi bridges, and protocols to evade detection. Advanced AI may automate the final cash‑out stage by creating fake, KYC‑compliant exchange accounts, and authorities shut domains linked to a major scam compound in Myanmar in December, the report notes.
