Fraudsters are increasingly targeting money itself – and AI is making it easier than ever for them, according to new research from Riskified.
New financial platforms for cryptocurrency, remittances, and precious metals have made moving money faster and more convenient. But convenience comes with risk.
Unlike traditional fraud, where criminals stole goods, today’s scammers go straight for cash, crypto, or precious metals. The speed, liquidity, global reach, and novelty of these platforms make them both attractive and hard to protect.
From opportunistic individuals to organised criminal networks, fraudsters are exploiting these markets in increasingly sophisticated ways. Many still use phishing and social engineering to trick victims into sending money, while others bypass humans entirely – using account takeovers, verification exploits, and gaps in know-your-customer (KYC) processes to steal funds directly.
There are roughly 10 000 cryptocurrencies today, with Bitcoin and Ethereum making up nearly 75% of the $3-trillion market. Global remittances are expected to hit $913-billion in 2025, and the precious metals market could top $500-billion by 2032.
Fraud is keeping pace: in 2024 alone, $4,6-billion was stolen in crypto, much of it via AI-generated deepfakes. In Japan, online fraud losses continue to rise with victims tricked into downloading fake apps or visiting fraudulent websites impersonating delivery and shopping services.
Fraud using crypto, electronic money, and prepaid cards has also been confirmed. Deloitte projects U.S. financial services could lose as much as $40-billion to fraud by 2027, much of it driven by AI. Fraud has now become a full-blown industry.
Riskified analysts have identified a growing dark web ecosystem that supports fraud at any scale. This includes voice-cloning apps, fraud-for-hire services, and lead lists of high-value targets. AI lets criminals create face-swaps, voice clones, fake selfies, and doctored handwriting. All to bypass KYC checks and access accounts, digital wallets, and crypto.
Financial platforms face rising risk as AI and industrial-scale fraud intersect. Protecting customers while keeping business moving requires agile, adaptive fraud prevention. First-time customers are especially challenging: they can be 40x riskier than users with a 30-day history. Fraudsters often play the long game, building trust before striking. Market volatility, currency swings, and crypto booms create additional openings for high-speed, low-value fraud.
KYC-verified accounts are highly prized, offering anonymity and legitimacy. Fraud-as-a-service providers now make it easier to bypass KYC, registering fake accounts, passing selfie verification, and submitting doctored documents using AI. Cash-out strategies include hijacking older accounts or moving funds through new accounts linked to stolen payment methods. Often, account holders don’t even know their credentials have been compromised.
To stay ahead, platforms need strong anomaly detection, identity clustering, and realtime decision-making.
Fraudsters often mimic legitimate behavior, so spotting masking tactics like proxies is crucial. Robust identity clusters, machine learning, and cross-platform data helps seperate real customers from fraudsters.