AI Output Disclaimer — placement guide

Why this matters (risk R-L6)

Attestlane uses third-party AI (Groq / Llama) to generate policies, power the advisor chat, and auto-fill security questionnaires. If AI output is wrong and a customer relies on it — especially when sending an auto-filled questionnaire to their own prospect — Attestlane could be pulled into a misrepresentation claim. A visible, consistent disclaimer is the cheapest mitigation.

Short inline disclaimer (use near every AI surface)

AI-generated. Review before use. This content was produced by an AI model and may be inaccurate or incomplete. It is not legal, audit, or compliance advice. You are responsible for verifying it before relying on or sharing it.

Where to place it (UI)

| Feature | Placement | |---|---| | AI Policy Generator | Banner above generated policy + footer line in exported policy files | | AI Compliance Advisor (chat) | Persistent one-line notice under the chat input | | Questionnaire Auto-Fill | Confirmation step before any answer set can be exported/sent — require an explicit "I have reviewed these answers" acknowledgement | | Exported documents (HTML/Markdown/PDF) | Footer line on every AI-generated export |

Stronger gate for questionnaire autofill (recommended)

Because auto-filled questionnaires leave your platform and go to a third party, require an explicit checkbox acknowledgement before export:

Export footer line (plain text)

Generated with Attestlane AI assistance. Review for accuracy before use.
This is not legal, audit, or compliance advice.

Implementation note: these are UI/copy changes. Per current guardrails the CTO will not modify frontend UI without founder approval — flagged here for sign-off.