# Disclaimers ## Description **Disclaimers** let you present important information to learners **before they start chatting with an agent**. You can require an explicit **User Agreement** (blocking modal that must be accepted to proceed) and/or display a **non-blocking Advisory** (informational notice). Copy is fully editable (**Markdown supported**), allowing administrators to align the language with **institutional policies**. --- ## Target Audience **Superadministrator · Administrator** --- ## Features #### Two Disclaimer Types - **User Agreement** – Blocks chat until the learner clicks **Accept**. - **Advisory** – Non-blocking banner/modal (e.g., *“AI can make mistakes.”*) #### Pre-Chat Enforcement When a **User Agreement** is enabled, the agent will not allow chatting until it’s accepted. #### Markdown Support Write disclaimers in **Markdown** for headings, links, and emphasis. #### Acceptance Confirmation After acceptance, a confirmation message appears in the chat. Any message a learner typed before accepting is queued and then answered automatically once accepted. #### Editable at Any Time Use **Edit** to update the text; defaults are provided for quick setup. --- ## How to Use (step by step) #### Open Disclaimers - In the agent header, open the dropdown → **Settings → Disclaimers** #### Add Content - Choose the disclaimer type(s) you want to show: - **User Agreement** (blocking) - **Advisory** (non-blocking) - Click **Edit** and enter your copy (Markdown supported) **Example (User Agreement):** - **Title:** “Course Agent User Agreement” - **Body (Markdown):** > “By continuing, you acknowledge that the agent may make mistakes and that you will verify important information with official course materials or staff.” #### Save & Enable - Click **Save** to apply changes - Ensure the **User Agreement** and/or **Advisory** toggles are **On** #### Test the Flow - Start a new chat with the agent - Verify behavior: - With **User Agreement** enabled, a modal appears and must be accepted to proceed - You cannot click out of it - On **Accept**, a confirmation appears and any message typed pre-acceptance is submitted and answered - With **Advisory** enabled, the notice appears but does not block chat #### Iterate as Needed - Revisit **Disclaimers** to refine wording, add links to policies, or switch between **Advisory** and **User Agreement** depending on the use case --- ## Pedagogical Use Cases #### Academic Integrity & Policy Acknowledgment Require acceptance of **honor-code reminders** or **collaboration rules** before users use the agent. #### Scope & Limitations Present an **Advisory** that clarifies the agent’s boundaries (e.g., *“No medical/financial advice; verify critical info.”*) #### Accessibility & Support Link to **disability services**, **office hours**, or **help desk resources** in a brief pre-chat notice. #### Sensitive Topics Use a blocking **User Agreement** to ensure users acknowledge **content warnings** or **research-ethics guidance**. --- By configuring **Disclaimers**, administrators keep Agentic OS use **safe, transparent, and aligned** with institutional expectations—while preserving a smooth chat experience once learners accept.