AI-driven fraud far more profitable, Interpol warns

Interpol's latest annual financial fraud report warns that AI-assisted fraud schemes are 4.5 times more profitable than non-AI ones. Criminals use generative AI to craft convincing phishing messages, eliminate language errors, and create voice clones from as little as 10 seconds of audio. Deepfake-as-a-service kits are available on dark web marketplaces at low cost, lowering the barrier to entry for sophisticated fraud. Interpol also flags the rise of AI-powered sextortion and the global expansion of scam centers beyond Southeast Asia into Africa, Latin America, and Europe — many staffed by trafficked individuals. Global financial fraud losses in 2025 reached an estimated $442 billion, a figure expected to grow as agentic AI matures and potentially automates more of the attack lifecycle.