AI for Nonprofit Fundraising: Six Workflows, Where AI Helps, and Where It Doesn't

AI is useful in fundraising where the work is preparation: drafting, sorting, summarizing and planning. It's not useful, and can do harm, where the work is relationship and judgment: deciding whom to ask, for how much, and how to say thank you so a donor believes it. The six workflows below keep that line in place.

I wrote a short book on this, The AI Fundraiser Playbook. This article covers the same ground in brief, with the rules we use at Storyline, the mission organization I lead. The book is $9.99 on Kindle as of October 2026, and I earn from it; you don't need it to use anything below.

Start at your level

Most organizations try to do too much at once. A simple way to think about it:

• First base: one person, one assistant. A staff member uses a chat assistant on an organizational account for drafts and summaries. Most small nonprofits should start here.
• Second base: an agent for one job. A set of instructions that does one repeating task, such as preparing a weekly draft for a person to approve.
• Third base: connected data. AI working with your donor database and email tool together. This needs clean data, careful permissions, and someone technical.

Each step needs the one before it. If staff aren't yet using AI well for drafts, connecting it to your donor database won't help.

1. A one-page donor brief before a meeting

Before meeting a donor, a development director gathers giving history, past conversations, the donor's interests and any public information. AI can turn your notes and the donor's public profile into a one-page brief: what they've supported, what they've said, what to ask about.

Where AI helps: organizing and summarizing what you already have. Where it doesn't: it can't know what isn't written down, and it may state guesses as facts. Mark anything from public sources with where it came from, and never put information into an AI tool your organization hasn't approved for donor data.

2. Fundraising emails that don't sound like AI

AI is good at a first draft: structure, a clear ask, a subject line or three to choose from. It's poor at the parts donors respond to, which are specific: a real person helped, a real number, a real thank-you from someone they know.

The rule we use: AI drafts, people send. At Storyline, agents can build Mailchimp drafts, and a check that runs before every command stops any agent from sending, scheduling or test-sending. I asked for that on October 20, 2026, because sending to supporters is a judgment I want a person to make. And never let AI invent a story, a quote or a result. If you don't have the story, write without one.

3. Sorting a messy donor list

Most small nonprofits have a donor list with duplicates, inconsistent fields and notes in the wrong column. AI can help propose groups, such as first-time donors this year, monthly donors, people who gave to a specific campaign, and draft the rules for each group, which a person then applies in the donor database.

Keep the actual records in your database. Ask AI to help write the rules and the filters, not to hold the list.

4. Re-engaging lapsed donors

AI can help you plan the outreach: which lapsed donors to contact first, based on rules you set, and a short sequence of messages tailored to what each group supported. A person decides the final list and writes or approves every message.

Watch your list hygiene too. On Storyline's website email sign-up, about 40 of 47 new contacts after October 13, 2026 turned out to be bots. If you're sending to people who never really signed up, no amount of good writing will help.

5. Preparing for a major donor meeting

For a major donor meeting, AI can draft the agenda, a short summary of the project you'll discuss, likely questions and plain answers, and a follow-up email for after. A person checks every number and every promise. A major gift conversation is the last place for a confident wrong fact.

6. Thank-yous

AI can draft thank-you letters and help you send them faster. Speed matters. But a thank-you has to be true: what the gift did, in words the donor will recognize. Use AI to get the draft started, then make it specific.

A 90-day plan

Days	Do this
1 to 30	Set up one assistant on your organization's account, at nonprofit pricing. Write a one-page AI policy. Use AI for first drafts of appeals and thank-yous, with a person editing every one.
31 to 60	Add the donor brief before meetings and the lapsed-donor plan. Count the hours saved and the corrections needed.
61 to 90	Pick the one job that saved the most time and turn it into a repeatable set of instructions (second base). Decide whether connecting to your donor data (third base) is worth the effort.

The tools are inexpensive. As of October 2026, ChatGPT Business and Claude Team are $8 a user a month for verified nonprofits, and the Gemini app is free with Google Workspace for Nonprofits; see our AI tools discount comparison for sources and eligibility. The nonprofit AI policy template covers the data rules.

For the bigger picture of agents in nonprofit work, read AI agents for nonprofits. And before adding any new tool, the free audit checks what you already pay against nonprofit pricing.