Use AI for first drafts, summaries and repeat writing tasks, never for decisions about people or for anything that contains donor or client data you haven't cleared. Write down three things on one page: what may go into AI, what never may, and who checks the output before it goes public. That page is what separates useful AI from scattered AI.
Is your nonprofit already using AI?
Almost certainly. In Virtuous's 2026 Nonprofit AI Adoption Report, which surveyed 346 nonprofits in December 2025, 92% were using AI in some way, but only 7% reported a major impact on their work, and 47% had no AI policy at all. In the UK, early findings from the 2026 Charity Digital Skills Report show little difference left between the adoption rates of small and large charities.
So the question for a small team isn't whether to use AI. It's how to use it on purpose.
Five uses that save a small team real time
- First drafts of appeals, emails and posts, from your own notes and past examples. You edit; the AI types.
- Turning one piece into many: one story becomes an email, three social posts and a short website update.
- Summaries: long grant guidelines, reports or meeting notes, turned into one page of what matters.
- Thank-you and update templates you personalize in a minute instead of writing from scratch.
- Checking your own writing: clarity, length, a reading level that suits your donors.
Five things never to put into AI
- Donors' or clients' personal details: names with addresses, health, finances, anything identifying, unless your tool and policy clearly allow it.
- Passwords, bank details or any login.
- Confidential board or legal matters.
- Children's personal information or photos.
- Anything you'd be uncomfortable seeing published.
The one-page AI policy
Copy this, adapt it, and have your board or director approve it.
[ORGANIZATION] AI use policy (one page)
1. Approved tools: [names of the AI tools staff may use, on organization accounts].
2. What may go in: public information, our own published content, anonymized notes.
3. What never goes in: personal data about donors, clients, volunteers or children; passwords; financial or legal documents; confidential board matters.
4. Human review: a named person reads every AI-assisted text before it reaches donors or the public. AI never sends anything on its own.
5. Honesty: we never use AI to invent stories, quotes, photos or results. Every story we publish is true and has permission.
6. Images: we don't use AI images of people or animals we serve as if they were real.
7. Questions go to: [name].
Approved by [name] on [date]. Reviewed every 12 months.
What about AI-generated stories and photos?
Don't. Donors give because the story is true. An invented beneficiary or an AI photo presented as real can cost you trust that took years to build, and in some places it can break fundraising rules. Use AI to tell your true stories better, not to make new ones.
Frequently asked questions
Which AI tool should a small nonprofit use?
The one your team will actually use every week, on a paid organizational account so you control the data settings. The policy matters more than the brand.
Will donors mind if we use AI?
They mind dishonesty, not tools. Use AI for drafting, keep a human in charge, and never fake a story or a photo.
How much time can AI save?
It depends on how organized your work is. Teams with written processes and saved templates save the most.
How ready is your organization? Take the free 2-minute AI Readiness Check. Take the check →