Copilot data, privacy and security
Microsoft's edit ·
Microsoft's commit message in MicrosoftDocs/microsoft-365-docs · commit a79eeb2 · +5 −1 lines · also on GitHub
⋯ 15 unchanged lines
1616
- trust-pod
1717
hideEdit: true
1818
ms.update-cycle: 180-days
19
−ms.date: 08/22/2025
19
+ms.date: 09/05/2025
2020
ms.custom: [copilot-learning-hub]
2121
---
2222
⋯ 168 unchanged lines
191191
### Does Copilot block prompt injections (jailbreak attacks)?
192192
193193
Jailbreak attacks are prompts designed to bypass Copilot's safeguards or induce non-compliant behavior. Microsoft 365 Copilot helps mitigate these attacks by using proprietary jailbreak and cross-prompt injection attack (XPIA) classifiers. These classifiers analyze inputs to the Copilot service and help block high-risk prompts prior to model execution.
194
+
195
+### What happens when foundation model changes occur?
196
+
197
+The AI models that power Microsoft 365 Copilot are regularly updated and enhanced. Model updates bring performance improvements, more advanced reasoning, and expanded capabilities, but they don't change your security, privacy, or compliance settings. For more information, see [Microsoft 365 Blog: Understanding foundation model changes in Microsoft 365 Copilot](https://techcommunity.microsoft.com/blog/microsoft_365blog/understanding-foundation-model-changes-in-microsoft-365-copilot/4440322).
194198
195199
### Committed to responsible AI
196200
⋯ 11 unchanged lines
Microsoft's Markdown source from MicrosoftDocs/microsoft-365-docs, © Microsoft Corporation, under CC BY 4.0. Changed lines with up to 3 unchanged lines around each; ChangeIntel kept this copy 10 Oct 05:49 UTC. The commit date is when the source changed, which can be hours or days before Learn published it.