Abliterated LLM finds $3,300 in live ElevenLabs keys; OpenAI/Anthropic keys already drained
deedydas · x · 2026-10-27
deedydas reports using an abliterated LLM to scan the internet for leaked API keys, uncovering $3,300 in live ElevenLabs credentials — 321 working keys with a rotation script tracking balance and cooldowns.
Notably, every OpenAI and Anthropic key it found was already drained, suggesting those providers may be actively monitoring and sweeping leaked keys.
More from Safety
- CA eMudhra blocks curl/wget User-Agents from CRL endpoints, Mozilla bug filed — kingslyj · 2026-10-27
- 300 Agents, 60 People: Runlayer Found Engineer Details Real-World Agent Security Failures — AI Engineer · 2026-10-27
- Tailscale confirms serious Tailworm SSH flaws; fix due within 24 hours — AIFlow_ML · 2026-10-27
- OpenAI details how it locked down frontier training environments and red-teamed them with its own agents — AxSaucedo · 2026-10-27
- GMO hack: account compromises real, ransomware encryption still unconfirmed — jd_pressman · 2026-10-27
- Inference.net serves refusal-removed DeepSeek V4.1 Flash Cyber at cost, KYC required — AccBalanced · 2026-10-27