# Anthropic takes risky AI tests offline after models breach govt sites

> **Open Intelligence Dossier** · First detected: 2026-10-13 21:40 UTC · Category: Business

## Multi-Source Evidence Table
| Source Outlet | Headline | Verification URL |
|---|---|---|
| BleepingComputer | Anthropic takes risky AI tests offline after models breach govt sites | [Source Link](https://news.google.com/rss/articles/CBMiuAFBVV95cUxNMmZCZTBvUm9YVUZPNnQ4TmRfSF95MFVJV00zc05YWnRHZ3hkUzBsM3FPSEhkLUxPaEJjVF9Rd0FwYWJxRmtOLTdTNVB1dmpIT3RCNHdlOUJTalY3eHRyZGE4NVZaUzZWR3ZFSlkwNDZEcW5IOVJaRlZmb2RtbG95VlZoQVBwSTlJczBBbG9PVjhSdmpTbFJiU3JaRmZWYml6U2RaaHZSaEJjQ3JhZkRoNmhXaTFFbm5Y0gG-AUFVX3lxTE5oVm8zWHlNNUtSTXRGejRpNjBEZXFiWFJlU1drNEtCNDREWG5GZ1F2ZkVZN3JtZVFjeUJyZTJmQ2JMbExzOGx4bjR5RUpnbHhUVkxTZG93ZFlxMUVfZnZIT2lPb0J3LTByc04yalhSdno2WTlHR1ZVbjFuODhZUDlPam85RG1VNlBxR3dfOEQ5NGVSR3hGOEJYSjJ) |

---
*Canonical Source: https://pulse.byoviral.com/trend/2026-10-13/anthropic-takes-risky-ai-tests-offline-after-models-breach-govt-sites*