The cryptocurrency sector faces increasing threats from AI-driven hacking, as recent incidents highlight vulnerabilities beyond traditional smart-contract exploits. Notably, a North Korea-linked group siphoned $285 million from Drift Protocol through a prolonged social-engineering campaign, while Kelp DAO lost $292 million due to a single-verifier flaw. Additionally, Humanity Protocol suffered a $30 million loss from a private-key compromise. These breaches underscore the persistent risks of social engineering, exposed keys, and human error, which AI can exploit by accelerating reconnaissance and identifying operational weaknesses. Security experts emphasize the need for robust defenses, including securing private keys and ensuring users have trusted interfaces to verify transactions.