AI, Impersonations Drove Crypto Scam Losses to Record $17 Billion in 2025: Chainalysis
In 2025, cryptocurrency scam losses surged to a record $17 billion, driven by increasingly sophisticated AI-enabled impersonation tactics, according to Chainalysis. The average scam payment rose by 253% to $2,764, with AI-linked schemes generating $3.2 million per operation—over four times more than non-AI scams. Scammers used deepfakes, face-swap tools, and AI-powered models—often sourced from Chinese vendors via Telegram—to convincingly impersonate authorities, causing government impersonation scams to increase 1,400%. Large-scale phishing, like fake U.S. “E-ZPass” alerts from China-based groups, also grew, while pig-butchering scams remained especially lucrative. Criminals increasingly avoided centralized exchanges, moving stolen funds through decentralized platforms (DEXs, DeFi bridges) to evade detection. Further, advanced AI aids not just the scams themselves, but the mass creation of fake accounts to cash out proceeds. Scam operations have centralized in parts of Southeast Asia, where human trafficking and forced labor fuel massive “pig butchering” compounds. Chainalysis highlights the dual impact: massive financial losses for victims and human rights abuses against trafficked scam workers.
