What is a honeypot endpoint?
A honeypot endpoint is a fake but realistic service, such as a decoy API or AI model endpoint, set up so that anyone who uses it reveals themselves and their methods.
What is a honeypot endpoint?
A honeypot endpoint is a fake but realistic service, such as a decoy API or AI model endpoint, set up so that anyone who uses it reveals themselves and their methods.
Why it matters
Attackers increasingly run AI agents that probe for exposed services and call whatever endpoint they are pointed at. A decoy that looks like a real model API or admin interface can capture an attacking agent's prompts, tools and goals, giving defenders early warning and indicators to block.
Example
Amazon said on 13 November 2026 that its MadPot decoys posing as AI model endpoints received attacker ARTEX agents' prompts, tool histories and findings between 3 October and 5 November, exposing AI-key theft and deepfake identity-check attempts (AWS).
How to defend against it
- Pin your own agents to approved model endpoints so they cannot be redirected to someone else's decoy.
- Deploy decoy API keys and endpoints in places attackers search, and alert on any use.
- Feed captured indicators (IPs, prompts, tool names) into detection rules.
- Never put real data behind a decoy; keep it isolated from production.
Related terms
Deception technology, canary token, LLMjacking
Sources
From AgentSec Brief, the daily brief on securing agentic AI. Subscribe free at https://www.agentsecbrief.com