# Microsoft says AI agents are like insider threats as ‘frontier models can’t be trusted’

- Source: TechCrunch
- Published: 2026-10-19T22:04:28+00:00
- Category: Tech
- Original: https://techcrunch.com/2026/10/19/microsoft-says-ai-agents-are-like-insider-threats-as-frontier-models-cant-be-trusted/
- Canonical: https://groundlevelnews.com/story/microsoft-says-ai-agents-are-like-insider-threats-as-frontier-models-c-b28c1a/

Microsoft wants enterprises to closely watch the agents employees use with systems not controlled by AI labs like OpenAI and Anthropic.

_Ground Level News shows a short summary only; the full article and all rights belong to TechCrunch._