End-User Privacy
What it means when an agent has End-User Privacy enabled.
When End-User Privacy is turned on for an agent, your conversation content and execution details are kept private — visible only to you.
End-User Privacy is not enabled by default. Administrators can choose which agents have this setting enabled.
What stays private
- Your messages and responses — the content of your conversations with the agent is not visible to administrators or other users
- Execution details — the step-by-step details of how the agent processed your request are hidden from others
What administrators can still see
Administrators retain access to metadata about End-User conversations (such as timestamps, usage counts, and other non-content information), but they cannot view the actual messages or responses.
How it affects your experience
- Your conversation history remains accessible to you as normal
- The agent works exactly the same way — End-User Privacy does not impact functionality, only visibility
A note on delegate agents
Some agents are built using other agents internally (called delegate agents). When End-User Privacy is enabled on an agent, it is automatically applied to all of its delegate agents as well. This ensures your data remains protected throughout the entire processing chain — not just at the surface level.