AI-linked crypto scams took in 4.5 times more per operation

Every AI trading tool marketed to retail investors shares the internet with AI tools sold to scammers. Chainalysis, the blockchain analytics firm used by law enforcement, quantified the second group in the scams chapter of its 2026 Crypto Crime Report, published January 13, 2026.

How it measured AI use

Most scammers pay for AI tools through ordinary channels, which leave no trace on-chain. But some buy from vendors, often Chinese-language sellers on Telegram, who take crypto for face-swap software, deepfake tools and large language models. Chainalysis flagged scam operations with on-chain payments to those vendors and compared them with scam operations without such links.

The gap

On average, scams with on-chain links to AI vendors extracted $3.2 million per operation, against $719,000 for those without, which Chainalysis describes as 4.5 times more revenue per scam. Median daily revenue was $4,838 against $518, and average transfers per day were 35.1 against 3.89, about nine times more transaction activity. Chainalysis also reports that 76% of AI-enabled scams fall in its high-value, high-volume quadrant nContext: n: scam operations not disclosed · calendar 2025 · Chainalysis 2026 Crypto Crime Report · margin: not disclosed.

Chainalysis reads the transfer count as a sign that AI lets scammers manage more victims at once, and the higher revenue as a sign that AI makes scams more persuasive.

The bigger picture

Chainalysis estimates crypto scams received at least $14 billion on-chain in 2025 and projects the figure could exceed $17 billion as more illicit addresses are identified. It reports that impersonation scams grew more than 1,400% year over year, and that the average scam payment rose to $2,764 in 2025 from $782 in 2024, an increase of 253% nContext: n: scam operations not disclosed · calendar 2025 · Chainalysis 2026 Crypto Crime Report · margin: not disclosed. Phishing kits sold as a service, some for under $500, made scams using them far more effective in dollar terms, according to the report.

Where AI trading meets AI fraud

For traders, the overlap is concrete. The fastest-growing categories in the report, impersonation and investment scams, are the ones that use fake support agents, fake advisors and deepfake endorsements. Pitches for 'AI trading bots' that promise returns are a common wrapper. Chainalysis cites a J.P. Morgan report on scammers using deepfakes and AI-generated content in romance and investment scams.

The industrialization goes beyond AI. Chainalysis reports that scams using phishing-kit services were 688 times more effective in dollar terms than regular scams, and those buying bulk social media accounts 238 times more effective. One kit vendor, Lighthouse, received more than 7,000 deposits and about $1.5 million in crypto over three years, according to the report, citing Cisco Talos.

The defenses Chainalysis suggests

The report calls for real-time fraud detection at exchanges and financial institutions, faster cross-border freezing of funds, and support for law enforcement in low-capacity jurisdictions. Its own data also shows the limits of tracing: AI tools bought with cards or bank transfers stay invisible, so the AI-linked group is a subset of AI-enabled fraud, not all of it.

What to take from it

The 4.5 times figure is a correlation drawn from scams that left an on-chain trail to AI vendors. Chainalysis does not claim AI alone caused the gap; larger, better-organized operations may be more likely to buy such tools in the first place. The direction is still clear: the cost of a convincing fake has dropped, and the payoff has risen.