Most AI policies end up as a PDF nobody opens. Staff hit a question mid-task — can I put this in Copilot? — and either guess or wait two days for an answer. This is the thing in between.
Yes. Drafting a newsletter is writing assistance, which is an approved use, and Copilot is the Trust's primary approved tool. Newsletters are published material, so this is fine — just make sure a person reviews the draft before it goes out.
Per: AI Use Policy · Approved AI Tools List
AI meeting transcription is approved for internal meetings only, but explicitly not for supervision or one-to-one meetings. Supervision often touches on personnel matters, which is why it's excluded.
Per: Approved AI Tools List
Please contact Alex Rivera, Operations Director, before proceeding.
The second rule lives in the approved tools list rather than the policy itself. The assistant finds it, says which document it came from, and stops rather than guessing.
Try asking it:
- “Can I paste a client's case notes into ChatGPT to summarise them?”
- “Can I use AI to help score job applications?”
- “Can I use AI to translate our housing leaflet into Polish?”
Look inside the admin side
Whoever owns the policy gets a dashboard: every question asked, filterable by verdict and date, with the documents editable in the browser. Log in and look around — it's a view-only account, so nothing you click will change anything.
aipolicy.bridgesstrategy.com/demo/admin/login
[email protected] · SeeTheDashboard
The log fills with whatever visitors to this page have asked, which is a fair illustration of the real thing: a pile of unfiltered questions, some of them revealing. That is what the policy owner reads once a month.
Why the answers hold up
- It only knows your documents. No general knowledge, no advice from the internet. If your policy doesn't cover it, it says so.
- A verdict, not an essay. Green, Yellow or Red, two sentences, and the document it came from.
- It knows when to stop. Anything ambiguous or prohibited ends with a named person to contact.
- It keeps a record. The log is a live map of where your policy is unclear, written by the people who have to follow it.
Two ways to build it
| Build it yourself | Purpose-built | |
|---|---|---|
| Setup | About fifteen minutes | About a day |
| Cost | Free, on seats you already pay for | Around $10 a month |
| Staff access | Anyone with a paid AI account | Anyone with the link — no login |
| Question log | None | Automatic and anonymous |
| Documents | Pasted into the assistant | Edited in a browser |
| Lives | Inside your AI tool | At your own web address |
Start with the free one. A policy your staff can actually question beats a better plan you haven't built — and you'll learn what people ask before spending anything.
Option 1 — build it yourself
Paste your policy and the prompt below into a custom assistant in whatever AI tool you already use. Staff ask it; anything unclear goes to a named person. You need your policy as text you can copy, and someone willing to be that contact.
Setup steps and the prompt
- Create a new custom assistant — a Custom GPT in ChatGPT, a Project in Claude, a Gem in Gemini, or an agent in Copilot.
- Copy the prompt below into its instructions.
- Replace the highlighted placeholders, and paste your policy where it says to.
- Save it and share the link — in Teams, on your intranet, or in an email signature.
You are the AI Policy Assistant for [ORGANISATION]. Answer only from the policy below. Never use outside knowledge about AI tools, vendors, or best practice — if it is not in the policy, you do not know it. Reply in this format every time: **[GREEN / YELLOW / RED]** Two or three sentences, plain language, explaining why. *Source: [section or heading you relied on]* GREEN = clearly allowed RED = clearly prohibited YELLOW = ambiguous, not addressed, or depends on specifics If the policy does not clearly cover the question, say so and answer YELLOW. Never fill a gap with general knowledge or a reasonable guess. End every YELLOW or RED answer with: "Please check with [NAME] at [EMAIL] before proceeding." Do not ask for or repeat real names, client records, donor details, or other personal information. If a question contains them, answer the general question and remind the person to leave personal details out. --- POLICY BEGINS --- [PASTE THE FULL POLICY TEXT HERE] --- POLICY ENDS ---
Paste the policy in rather than attaching it as a file. Attached files get searched in fragments, so the assistant sometimes misses a clause. Pasted text is always fully in view.
More than one document? Paste them all in, and add this line above them: “These sources are listed in order of authority. The AI Use Policy governs — if a supporting document conflicts with it, follow the policy and say you noticed the conflict.”
Option 2 — a purpose-built assistant
A small web application at your own address, which is what you tried at the top of this page. Staff need no login at all — which matters, because the people most likely to have a policy question are often the ones without a paid AI licence. Every question is logged automatically and anonymously, and whoever owns the policy edits the documents themselves.
What it takes to run one
Around $10 a month for hosting and AI usage combined, an account with an AI provider, and about a day to set up. It runs on infrastructure you own and pay for, so it stays yours — and everything exports as a single file if you ever move it.
I can build it with you or for you. If you have someone technical in-house, it is a genuinely small application: the version behind the demo on this page was built in an afternoon with an AI coding tool, which is rather the point.
Getting value from the question log
The answers help one person for a minute. The log is what improves the policy. Half an hour a month is enough:
- Read the month in one sitting. Themes are obvious in a batch and invisible one at a time.
- Look for repeats. The same question three times is a gap in the policy, not three confused people.
- Edit the document, not the assistant. Add the missing clause and the answers change immediately — and stay traceable to an approved document.
- Tell people what changed. Staff who see their question become policy ask more questions.
Don't let it learn from its own answers. A policy tool has to be auditable — you need to point at the clause behind any verdict. Keep the documents as the single source of truth.
Or we can do it together
I help nonprofits put this in place — from a single working session to a hosted version running on infrastructure you own, with your documents loaded and your team shown how to keep it current.
Stay current on AI in the nonprofit sector
A weekly digest of curated news and use cases for nonprofit leaders. Sent every Monday.
No spam. Unsubscribe anytime. We do not share subscriber data.