AI-powered crypto scams surged in 2025 and early 2026, with a 456% rise in gen-AI-enabled scams between May 2024 and April 2025, and roughly 60% of deposits into scam wallets now linked to AI tools, illustrating widespread fraudster adoption of AI.

Scammers are leveraging AI to craft highly realistic, personalized content across channels, including phishing emails, impersonation via deepfakes and voice cloning, and tailored messages that dodge traditional filters and detection systems.

Common AI-driven scam types now include deepfake audio and video impersonations, AI-generated phishing, fake AI trading platforms and bots, real-time AI-enabled voice cloning, pig-butchering scams augmented by AI, prompt-injection attacks on agentic browsers and wallet-connected assistants, AI-generated IDs to bypass KYC, and social botnets on platforms like X.

AI enables attackers to conduct massive, rapid campaigns across email, Telegram, Discord, SMS and social media, localizing content to victims to increase believability and scalability.

Defense recommendations emphasize using 2FA or passkeys, avoiding unsolicited links, treating seed phrases as offline private keys, using official support channels, and leveraging hardware wallets, with ongoing education and security practices underscoring the defense.

Crypto remains a prime target due to fast, irreversible transactions, around-the-clock markets, and a global audience, with scammers playing on emotions like greed and fear.

A standout real-world case involved a $25 million deepfake-enabled executive impersonation in Hong Kong, underscoring how convincing these schemes can be.