British figures tell the story plainly: deepfake attempts there jumped 94% in a year while total fraud volume stayed nearly flat. The volume is steady, but each attempt is sharper, more personal, and increasingly run by machines that need little human input, all of which works in the criminal’s favor. The category covering advanced deception climbed 180% year over year, drawing on intricate trickery, social manipulation, and AI-built identities designed to walk straight past detection systems. Most AI scams build on synthetic identity fraud and deepfakes. Agentic AI and automated fraud campaigns to grow. AI-mutated malware and adaptive cyberattacks to grow. Vishing with AI (exploiting trust in people you know) to grow. Across 2025 and 2026, the five hardest hit were dating, online media, financial services, crypto, and professional services, with dating and online media each at a 6.3% fraud rate. Banking faces a rising tide of deepfakes and AI-generated paperwork that wears down old verification, leaving the industry little choice but to adopt sharper detection or gamble with customers and system integrity. Payment fraud breaks into categories.