Services — Beta

PolicyBuilder

A short structured interview produces a first-draft AI use policy for your organisation — grounded in EU AI Act, UK Government AI Playbook, and CSA AICM methodology. This is a beta feature: expect rough edges, and treat the output as a starting draft, not a finished, legally-reviewed document.

Before you start — a note on privacy

PolicyBuilder runs on Google Gemma 4 (31B) via Google’s free AI tier. On that tier, Google may use what you type to train its models, and human reviewers may read it. So please:

  • Describe your organisation in general terms — no real names, contact details, health, safeguarding, or membership records.
  • Use placeholders like “[Name]” or “our safeguarding lead”, and add the real detail yourself once you download the draft.
  • Mantle stores none of this conversation, but each reply is generated by passing your text to Google.