AI agents are set to power crypto payments, but a hidden flaw could expose wallets
Summary
A research team from institutions including the University of California and Fuzzland has identified a critical security risk within "LLM routers," which act as intermediaries for AI services. These routers can intercept sensitive data, including private keys and API credentials, allowing attackers to modify commands or exfiltrate secrets. With AI agents increasingly tasked with autonomous financial transactions, this vulnerability poses a severe threat, as demonstrated by the theft of a $500,000 wallet and the potential to compromise hundreds of hosts.
(Source:CoinDesk)