Update AI legal disclaimer to account for agentic functionality
Description
https://gitlab.com/gitlab-org/gitlab/-/issues/562399 introduced a new legal disclaimer requirement for agentic AI features. It is quite lengthy and deviates from our current guidance. Given all AI interactions are likely to have agentic elements we should standardize and align on guidance.
Examples
Current Pajamas guidance:
You must include a disclaimer to inform the user that the feature is powered by AI. This is a legal requirement in some regions and by certain third-party AI services.
Flag AI-generated content with the disclaimer by AI. For example: “Generated by AI” or “Summarized by AI”.
Close to that disclaimer, show a message to encourage the user to verify the AI-generated content. If space allows, show the following directly in the UI or in a popover: “Content generated by AI should be seen as a starting point and verified before use. It may be incorrect, inappropriate, or diverge from your organization’s standards.” Or, if space is a concern, just “Verify before use.”
Show the disclaimer only once per context, and preferably under the AI-generated content, in a way that is clear that it applies to all content within that context.
New legal disclaimer for agentic features:
GitLab Duo can autonomously change code. Responses and changes can be inaccurate. Review carefully.
Actions to take
1. Resolve open questions
We should work with Legal and Technical Writing to improve the disclaimer given it will be used frequently.
- Users can create custom agents and flow. Should we refer to "GitLab Duo" or is a generic "AI" more appropriate. Users can also run third party agents from Claude Code or Codex.
- The sentence is lengthy and will be hard to fit into the UI, esp. the Duo Chat panel. Could we shorten to something like "AI Generated" and then link to our full legal disclaimer. This is what some competitors do.
2. Update Pajamas guidance
This page will need updating: https://design.gitlab.com/patterns/ai-human-interaction/#disclaimer