AI for not-for-profits: where it genuinely helps, and what to get right first.

Most not-for-profits I work with have the same problem in a different shape. The demand for their services keeps growing, the funding doesn't, and the people doing the work spend a surprising share of their week on writing, reporting and admin rather than on the mission.

That is exactly the kind of work AI is good at. It is also why the sector has moved quickly. Infoxchange's 2025 Digital Technology in the Not-for-Profit Sector report found that around two-thirds of the organisations it surveyed were already using generative AI. The same report found that only 14% had an AI policy or guidelines in place.

So the opportunity is real, and so is the gap. Here is where I think AI earns its place in a not-for-profit, and what to sort out before you lean on it.

Where AI genuinely helps

Grant applications and acquittals

Grant writing is mostly reworking material you already have: your mission, your programs, last year's outcomes and the budget. AI can turn that into a solid first draft against a funder's criteria in minutes, and help you reshape it for the next funder. The same goes for acquittal reports. A person still needs to check every claim and every number, but starting from a draft rather than a blank page saves hours.

Meeting notes and board papers

Infoxchange found almost half of organisations using generative AI were using it for notetaking and summarising. It is an easy win. Transcribe a team or board meeting, get a summary with actions and owners, and spend the saved time following up. Board papers can be condensed into a one-page brief for directors who volunteer their time and have little of it.

Supporter and donor communications

Thank-you letters, newsletters, appeal emails and social posts all take time to write well. AI can draft them in your organisation's voice, personalise them for different supporter groups, and help a small team keep up a steady rhythm of communication that would otherwise slip.

Finding answers in your own documents

Policies, procedures, volunteer handbooks and program guidelines tend to live across SharePoint sites, shared drives and inboxes. Tools like Microsoft 365 Copilot can search across what staff already have access to and answer questions in plain English. For a new volunteer coordinator, "what's our process for a volunteer working with children?" answered in seconds, with a link to the source document, is a real help.

Impact reporting

Funders and boards want outcomes, not activity counts, and pulling those outcomes together from spreadsheets and case systems is slow. AI can help summarise the data, draft the narrative around it, and with tools like Power BI, turn it into a report that updates itself rather than one rebuilt by hand every quarter.

What to get right first

A short AI policy

It does not need to be long. One or two pages that say which AI tools staff and volunteers may use, what information must never go into them, and who to ask when unsure. Writing it forces the conversations that matter, and it gives your people permission to use AI properly rather than quietly.

Know where your data goes

Many not-for-profits hold information about people at their most vulnerable: health details, housing situations, family circumstances, case notes. Pasting that into a free consumer AI tool on a personal account is the risk to remove first. Business versions of the major tools, used with a work account, come with commitments about how your data is handled and whether it is used to train models. Check those terms for the tool you choose, and make it the only one your policy allows for anything sensitive.

Clean up who can see what

AI tools that search your documents respect your existing permissions. That sounds reassuring until you realise how many organisations have SharePoint sites and shared folders open far wider than intended. If a volunteer can technically open the HR folder today, Copilot can surface it for them tomorrow. Reviewing permissions before you switch it on is the most important preparation step. We have written more about getting ready for Microsoft 365 Copilot and finding overshared data.

Put it in front of the board

Under the ACNC Governance Standards, a charity's responsible people must act with reasonable care and diligence and must not misuse information they gain through their role. AI makes it easier to misuse information without meaning to, so boards should know which tools are in use, have AI on the risk register, and review the policy periodically. A five-minute standing item at board meetings is enough to start.

Get the security basics in place

AI sits on top of your existing systems, so it inherits their weaknesses. The same Infoxchange report found only 23% of not-for-profits had a documented cyber security plan, and that multi-factor authentication was far from universal, especially in volunteer-led organisations. If staff accounts aren't protected with MFA, an AI tool connected to all of your documents makes a compromised account much more damaging.

A note on licensing

Check what you are actually paying for before adding AI. Microsoft ended its free Microsoft 365 Business Premium grant for nonprofits in 2025 and moved to discounted pricing instead. Microsoft 365 Business Basic remains free for eligible organisations, and there is nonprofit pricing on both Business Premium and Microsoft 365 Copilot. Some organisations missed the change notices and were caught out. Before you budget for Copilot, confirm what you're licensed for today and what eligibility you have.

Where to start

1. Pick two repetitive tasks that eat staff time, such as grant drafts and meeting notes, and start there.
2. Write the one-page AI policy and share it with staff, volunteers and the board.
3. Check your Microsoft licensing and nonprofit eligibility.
4. Review permissions on SharePoint and shared drives before connecting any AI tool to them.
5. Run a short pilot with a small group for four to six weeks, and track the hours saved before rolling it out further.

We work with not-for-profit and community organisations on exactly this groundwork, from data classification frameworks for Jawun to a Microsoft 365 and SharePoint migration for Your Place Housing. If you'd like to know whether your organisation is ready for AI, our AI readiness assessment is a good place to start.