Honeypot of 2M fake-secret prompts finds three AI hosting providers already breached
ccerrato147 · x · 2026-11-21
Over five months, hdmoore's team at Dragos sent about two million prompts carrying fake secrets to popular AI services. Three hosting providers turned out to have been breached without knowing it — stolen AWS keys were used in tens of thousands of calls.
The same investigation surfaced the Prime Intellect compromise: customers requesting DeepSeek via OpenRouter got a different model and had no way to tell; OpenRouter notified them by email this week. Prime Intellect found no sign prompts were copied out but can't rule out that they were read.
Related event: 2 Million Canary Tokens Expose Systemic Leaks Across the AI Supply Chain(11 posts)→
More from Infra
- Microsoft's USB4 RDMA lets mini PCs pool memory to run Qwen3-235B — JFPuget · 2026-11-22
- LTX 2.5 keeps crashing on AMD 9070 XT with driver timeouts while MiniMax H3 runs fine — ZestycloseReading504 · 2026-11-22
- FT: AI server suppliers raise prices up to 45% months after deals are signed — nordicinst · 2026-11-22
- China Installed 59% of New Industrial Robots in 2026 Report, ~354,000 Units — ingliguori · 2026-11-22
- MLX over Vulkan hits 124 tok/s on RTX 5060 Ti, beating llama.cpp CUDA by 25% — SumitGup · 2026-11-22
- Intruder swapped Prime Intellect's hosted DeepSeek for an abliterated copy for 53 hours — mawcopolow · 2026-11-22