AI, Impersonations Drove Crypto Scam Losses to Record $17 Billion in 2025: Chainalysis

A new report from blockchain analytics firm Chainalysis reveals that global losses stemming from cryptocurrency-related scams hit an unprecedented high of $17 billion in 2025, marking a significant escalation driven primarily by the integration of advanced Artificial Intelligence (AI) and sophisticated digital impersonation techniques. This staggering figure represents an estimated 45% increase over the losses recorded in 2024.

Chainalysis highlights that generative AI tools have fundamentally altered the operational landscape for crypto fraudsters. AI minimized the operational complexity for large scam syndicates, enabling the rapid creation of hyper-realistic phishing websites, highly convincing social engineering scripts, and personalized fake identities at unparalleled scale. This technology dramatically lowered the barrier to entry for complex fraud schemes, making them accessible to a wider array of criminal organizations.

Crucially, the proliferation of deepfake technology allowed fraudsters to execute ‘trusted entity’ impersonation scams with alarming success. The report tracked a sharp rise in scams involving synthesized voice and video used to impersonate CEOs, government regulators, and professional financial advisors. These deepfake campaigns often targeted high-net-worth individuals and corporate treasuries, exploiting trust via fabricated video conferences or personalized voice messages urging urgent crypto transfers or investment in fraudulent platforms. Additionally, established schemes like “Pig Butchering” (romance/investment scams) became far more potent, as AI-generated personas maintained seamless, hyper-realistic, and psychologically manipulative communication over extended periods.

Of the $17 billion lost, investment scams accounted for approximately 60%, a category increasingly fueled by automated AI bots performing sophisticated rug pulls that were difficult to distinguish from legitimate market activities. However, the losses attributed directly to AI-driven impersonation and phishing tactics showed the fastest rate of growth, quadrupling in some jurisdictions. Furthermore, AI-powered transaction obfuscation tools and advanced mixing services used by these syndicates made the tracing and recovery of stolen assets significantly more challenging for law enforcement throughout 2025. Chainalysis urges the global crypto community and regulatory bodies to prioritize the development of AI detection tools and mandatory identity verification protocols to combat this technological arms race in cybercrime.

Source: AI, Impersonations Drove Crypto Scam Losses to Record $17 Billion in 2025: Chainalysis