Tag: LLM abuse

Hackers exploit misconfigured proxies to access paid LLM services

January 10, 2026

Cyber Security

Misconfigured proxies LLM abuse has emerged as a growing threat as attackers exploit weak network configurations to gain free access to paid large language model services. Instead of breaching AI providers directly, threat actors target poorly secured proxy servers that organisations use to route API requests. This method allows attackers to consume expensive AI resources…