What can AI do for a UK charity or non-profit?
Yes, and some of it is already being used by charities who've tested it on real donor volumes and grant applications. AI can draft grant applications from your impact data and project plans, personalise donor communications without your team writing hundreds of individual emails, and extract information from Gift Aid declarations without someone typing them manually. The question is what's ready now versus what's being sold by people who've never written a grant application or managed a donor database.
Charity work is a mission-driven business with limited resources and regulatory requirements. Your team spends substantial time on work that doesn't directly deliver your mission: writing grant applications, updating donors on impact, chasing Gift Aid declarations, coordinating volunteers, and reporting to the Charity Commission. The bit that actually achieves your charitable purpose (delivering services, supporting beneficiaries, running programmes) takes up far less time than the fundraising and administration that keeps the organisation running.
Where the time actually goes
Grant applications and trust fundraising. Every grant application needs the same information presented differently: what you do, who you help, what the project will achieve, how much it costs, how you'll measure impact, why you're the right organisation to deliver it. Your fundraising team reads each funder's guidelines, adapts your standard information to match what they're asking for, writes it in their preferred format, and submits it by the deadline. If you're applying to multiple trusts and foundations, that's the same core information rewritten dozens of times a year.
Donor communications and stewardship. Donors want to know their money made a difference. That means thank-you emails after donations, impact updates during campaigns, annual reports showing what you achieved, and personalised communications for major donors. Your team writes these updates, pulls together the impact data, formats it appropriately for different donor segments (regular givers get email updates, major donors get personal letters, corporate sponsors get detailed reports), and sends them out. Much of the information is the same, but each communication needs tailoring to the audience.
Gift Aid processing and HMRC claims. Donor makes a gift, you need their Gift Aid declaration to claim the extra 25%. That means collecting declarations (online forms, paper forms, phone declarations), checking they're complete and valid, entering the information into your database, matching donations to declarations, and submitting claims to HMRC. If a declaration is incomplete or the donor details don't match HMRC records, you have to chase them. It's administrative work that's essential for your income but doesn't feel like it advances your mission.
Volunteer coordination and communication. You've got volunteers helping with various activities (events, service delivery, fundraising, administration). Someone has to recruit them, match them to opportunities, communicate what's needed and when, send reminders, track their hours (for impact reporting or insurance), and thank them for their time. The coordination is constant, and if volunteers don't get clear communication or feel appreciated, they stop volunteering.
Three things AI can genuinely do here
1. Draft grant applications from your project plans and impact data
What it does: You're applying for a grant. You upload your project plan, your impact data from similar previous work, and the funder's application guidelines. AI reads the guidelines, identifies what information they're asking for (project objectives, beneficiary numbers, outcomes, budget breakdown, evaluation methods), pulls the relevant information from your documents, and drafts the application in the structure and word count they've specified. Your fundraising team reviews it, adds the specific details and examples that make it compelling, adjusts the tone to match the funder, and submits it.
Why it works for charities: Grant applications are mostly information assembly and reformatting. You know what your project will do, who it will help, and what it will cost. The funder wants that information in their specific format with their specific questions answered. AI can do that initial assembly faster than a person reading the guidelines and copying information from various documents. Your team spends their time on making the application compelling (the case study that illustrates impact, the explanation of why your approach works, the evidence that you can deliver), not on the mechanical work of filling in forms.
Roughly what it takes: This needs your project information and impact data to be reasonably well-documented and accessible (not scattered across emails and people's heads). If your project plans are vague or your impact data isn't captured systematically, you'll need to improve that first. Once you have decent documentation, this is a tool you can start using relatively quickly. You'll need to check the AI's output carefully (does it answer the actual question, does it use appropriate language for this funder), but it saves substantial time per application. And before any of it, the data questions below.
2. Personalise donor communications at scale, with honest automation
What it does: You're sending an impact update to your donor base. Your system segments them (regular givers, one-off donors, major donors, corporate sponsors), pulls the relevant impact data for each segment (regular givers get overall impact numbers, major donors get details of the specific project they funded), and generates personalised emails or letters. "Thanks to your monthly gift of £10, we were able to provide 47 counselling sessions to young people this quarter" versus "Your donation of £5,000 funded our youth mental health project, which supported 23 young people through one-to-one counselling." Your team reviews the outputs, adjusts anything that doesn't sound right, and sends them.
Why it works for charities: Donor communications are mostly combining standard information (what you achieved this period) with personal details (what they gave, when, what it funded). A human doesn't add much to that combination except the time it takes to write it. AI can generate hundreds of personalised updates faster than your team can write them individually, and donors get communication that feels relevant to them rather than generic. And because this is mostly automation with some AI for the personalisation, it's reliable and doesn't need constant retraining.
Roughly what it takes: This needs your donor database to have good information (donation history, preferences, which appeals or projects they've supported) and your impact data to be captured systematically. If your database is messy or your impact tracking is patchy, the personalisation won't work well. Fix your data quality first, then automate the communications. Start with simple segments (regular givers vs one-off donors) and expand as you get more sophisticated. And before any of it, the data questions below.
3. Extract information from Gift Aid declarations and match them to donations automatically
What it does: Donor completes a Gift Aid declaration (online form, paper form, phone call transcribed). AI extracts the key information (full name, address, confirmation they're a UK taxpayer, date), checks it's complete, matches it to their donation record in your database, and flags any issues (name doesn't match the donation, address is incomplete, declaration is ambiguous about tax status). Your team reviews the flagged items, fixes what can be fixed, and chases the donor for clarification where needed. Clean declarations get processed automatically and included in your next HMRC claim.
Why it works for charities: Gift Aid processing is data extraction and matching. The information you need is on the declaration, it just needs transferring to your database and checking against HMRC requirements. AI can do that extraction and matching faster and more accurately than someone typing manually, especially if you're processing high volumes. Your team spends their time on the exceptions (incomplete declarations, mismatches that need investigation) rather than on routine data entry.
Roughly what it takes: This is relatively straightforward if you're collecting declarations digitally (online forms). If you're still working with paper declarations, you'll need to scan them, which adds a step. You'll need to check the AI's accuracy carefully at first (does it correctly extract names with unusual spellings, does it catch incomplete addresses), but once it's reliable, it saves time and reduces errors that could cause HMRC to reject your claim. And before any of it, the data questions below.
Before any of this touches a donor's data
Everything above involves software reading grant applications, donor records and communications that carry personal data and, in some cases, information about people's circumstances that is sensitive in practice even where it is not special category data in law. Trustees are accountable for what happens to it.
The Fundraising Regulator published specific guidance on using AI in fundraising in December 2025. Its position is neutral on whether to use AI, but clear about the conditions: it must be legal, open, honest and respectful under the Code of Fundraising Practice; trustees must be able to scrutinise how it is being used, not leave it to one enthusiast in the organisation; a risk assessment proportionate to the use before it starts; the privacy policy must say how AI processes personal data for fundraising purposes; and a human checks the accuracy, fairness and legality of any AI-generated content before it is used in fundraising. That last point applies directly to opportunities 1 and 2 on this page.
The Charity Commission does not expect to publish AI-specific guidance and expects trustees to apply existing duties. A governance policy covering the use of AI tools (what they are used for, who approved them, and how they are reviewed) is in the spirit of what both bodies expect.
Gift Aid declarations (opportunity 3) are HMRC records rather than fundraising data, but the data still belongs to donors, and the ICO baseline applies.
Links: Fundraising Regulator, Guidance for using artificial intelligence in fundraising (December 2025); ICO, Artificial intelligence hub.
What it won't do
It won't replace the human understanding of what makes a compelling case to a funder. AI can draft a grant application that answers all the questions. It can't tell you which case study will resonate with this particular funder, or how to position your project to align with their current priorities, or what tone will work best. That's still a fundraiser's job, and it's what determines whether you get funded.
It won't fix poor impact measurement or vague project plans. If you don't have clear data on who you help and what difference you make, AI can't create it for you. If your project plans are vague or your outcomes aren't measurable, automated grant applications will just expose that weakness faster. The charities that get the most out of this are the ones who already capture impact data systematically. If your impact measurement is weak, sort that first.
It won't handle donor relationships that need personal attention. If your major donor wants a phone call to discuss how their funding was used, or your long-standing supporter has specific questions about your strategy, they need human contact from your team. Automated communications work for keeping large numbers of donors informed. The relationships that matter most to your income still need personal stewardship, and trying to automate those will damage them.
Worth noting: much of what gets sold as "AI" is actually straightforward automation, which is often better news (cheaper, more reliable, easier to maintain). If you're not sure what counts as AI and what doesn't, we've written a plain-English explanation that cuts through the marketing.
If you're not yet sure AI is worth looking at for your business at all, this is worth reading first.
Scan your business now
This page describes what works across the sector. Your charity is specific: different cause area, different funding mix, different bottlenecks.
Run a free scan and get three AI opportunities written for your organisation, based on your website and what you actually do. Takes about a minute, no sign-up required, and if the best answer isn't AI the report will say so.
Run a free scan for your business
Free, takes about a minute, no email needed to read your report.