Evidence · DiggingBeagle record
InferD: credential-theft attempts against exposed model-tool honeypots
First-party paper on instrumented AI-serving and MCP honeypots and attempted credential extraction using encoded requests.
- Published
- Oct 30, 2026
- Source role
- primary disclosure
Evidence record
First-party paper on instrumented AI-serving and MCP honeypots and attempted credential extraction using encoded requests.
Claim-level citations (1)
- supportsMCP honeypots received hundreds of attempted credential-theft payloads without verified execution: InferD reports 601 attempted encoded credential-theft requests against monitored AI-related service surfaces, without proving successful execution or key exfiltration.
InferD paper, honeypot deployment, request classification and result limits