AI interaction logs allow admins with the Enterprise Guard add-on to collect and review records of Miro AI usage across their organization. By enabling AI interaction logs, admins can provide security, compliance, and governance teams with greater visibility into how AI features are used and what information is processed by AI systems.
AI interaction logs help organizations:
- Monitor how AI features are used across the organization
- Support governance, compliance, and security reviews
- Provide visibility into information shared with AI systems
- Strengthen trust and responsible adoption of AI tools
AI interaction logs capture records of interactions between users and AI-powered features in Miro. These records help organizations review how AI is used across the organization and support internal auditing, governance, and compliance processes. When enabled, AI interaction logs capture:
- User prompts submitted to AI features and the final responses shown to users
- Prompts sent to the AI model and responses returned by the model during an interaction
- Tool invocation arguments and results returned by tools during an interaction
Limitations
- Not every tool call is captured yet. Tool results can be truncated, and truncated results are marked as such.
- Interactions from external AI agents that call Miro through the Miro MCP server aren't currently logged.
- Moderation events and sensitive prompt blocks aren't currently logged. Where the AI blocks or refuses a request, you may see a failed outcome, but not the check that caused it.
- Images and files aren't included in AI interaction logs. An entry can indicate that content was an image or a file, but the file itself isn't included.
- Miro's internal system instructions aren't included in the logged content.
For more information on how to enable AI interaction logs and configure log retention, see our Help Center documentation.