Your Nonprofit's AI Toolkit: What to Automate and What to Keep Human

Almost every nonprofit is using AI now. Very few can tell you what it changed.

In the 2026 Nonprofit AI Adoption Report from Virtuous and Fundraising.AI, 92% of the 346 nonprofits surveyed said they use AI tools in some way, yet only 7% reported major improvements in what their organization can do (NonprofitPro). Most of that use is one person, one prompt, one deadline at a time.

So the better question for a small communications team is where AI belongs in your week, and where it has no business being.

How can nonprofits use AI for marketing?

Nonprofits get the most from AI on repeatable, behind the scenes marketing work: first drafts, repurposing content, sorting data and reporting. The parts that carry trust, like stories about the people you serve, donor relationships and final approval of any appeal, need a person. Put that split in writing so your whole team follows the same rules.

Why isn't AI saving nonprofits more time yet?

Most teams picked up tools before they designed workflows. The same report found that 81% of nonprofits use AI individually and ad hoc, only 4% have documented, repeatable workflows, and nearly half have no formal AI policy (NonprofitPro).

Think of the family recipe nobody wrote down. Titi makes the best arroz con gandules on the block, but when she isn't in the kitchen, nobody can repeat it. A prompt that works beautifully for your development director lives the same way: in one head, on one laptop. When she moves on, the gain goes with her.

GivingTuesday's AI readiness research points in the same direction. Across 930 nonprofits surveyed in 2024, the strongest predictor of AI readiness was having a dedicated technology or data specialist on staff. Budget and headcount were weaker predictors (Chronicle of Philanthropy).

Which marketing tasks should nonprofits automate with AI?

Start where the work repeats and the risk is low. A good test: if a mistake would get caught in review and cost you ten minutes, AI can take the first pass.

• First drafts of newsletters, social captions and event recaps, built from notes a staff member wrote
• Repurposing one approved piece (an annual report, a speech, an impact story) into email, social and web versions
• Translating content into Spanish, followed by a review from a fluent staffer who knows your community
• Summarizing survey responses, program feedback and meeting notes into themes
• Cleaning and tagging CRM and email lists, and suggesting audience segments
• Monthly reporting dashboards and plain language summaries for your board
• Prospect research briefs that a development officer then checks

On translation, a tool can swap the words. It can't tell you whether "guagua" means a bus or a baby to the reader in front of you. In Puerto Rico it's the bus; in Chile it's the baby. Your community's Spanish deserves a human ear.

What should stay human in nonprofit storytelling?

Anything that asks someone to trust you.

Donors notice. In research from BBB's Wise Giving Alliance, 55% of participants said they would be discouraged from giving if they knew an appeal had not been checked for accuracy by a staff member. Among people with household incomes of $200,000 or more, 70% said an appeal with an AI generated image would discourage them from giving (Give.org).

A 2026 study in the Journal of Business Ethics found that telling Facebook users a charity's image was AI generated cut clicks by about 25%. Companies running the same disclosure saw no penalty. The fix was simple: when the charity explained that AI helped it cut costs and put more help toward the children it served, the negative effect disappeared (iMarketSmart summary of Zhao, Zhou and Chen).

Keep these in human hands:

• Stories about the people you serve. Interviews, quotes, consent and photos come from real people who agreed to share them.
• Images of your community. Use real photos with permission. An AI generated child in a fundraising appeal puts your credibility on the line.
• Major donor and partner relationships. AI can prepare the briefing; a person makes the call.
• Final approval. Every public piece gets a human read for accuracy, tone and cultural fit before it goes out.
• Sensitive data. Donor records, case notes and anything about a client's legal or health situation stay out of public AI tools. Among regular AI users in the Virtuous report, privacy and security was the top concern at 32% (NonprofitPro), and in Okta for Good's survey of nonprofit IT and operations leaders, 57% named governance and oversight as their top concern with AI agents (Okta).

How do we write an AI policy for a nonprofit marketing team?

One page is enough to start. Answer four questions:

1. Which tools are approved, and who pays for them?
2. What data never goes into an AI tool?
3. Which content needs a named human reviewer, and who is that person?
4. How will we tell donors and the community when we use AI, and why?

The fourth question is where many teams freeze. The research above gives you the script: say what AI does for you and what it frees your people to do. "We use AI to draft routine updates so our staff can spend more time with families" lands very differently from silence followed by a donor finding out on their own.

Then write down the prompts and steps that work, so they belong to the organization and survive staff turnover. That is the recipe card for Titi's arroz.

Key takeaways

• AI use in nonprofits is nearly universal, and major impact is rare. Workflow and policy explain the gap.
• Automate drafts, repurposing, data cleanup and reporting.
• Keep stories, images of real people, relationships and final approval human.
• Protect donor and client data with written rules.
• If you tell donors you use AI, explain the purpose. They respond to the why.

Where does CAMI Studio fit?

CAMI Studio works with mission driven organizations on marketing operations and impact storytelling: workflow automation, MarTech stack audits, CRM builds, and messaging that sounds like the community you serve. If you want to see where AI fits in your marketing before you buy another tool, start with CAMI OS, our diagnostic that suggests where to begin (no sales call needed), or explore our services.

FAQ

Should nonprofits tell donors when they use AI?

Yes, and say why. A 2026 study found that disclosure alone lowered engagement with a charity's appeal, while a short statement of purpose, such as using AI to cut costs and serve more people, removed that penalty.

Can AI write grant proposals for a nonprofit?

AI can help organize a draft, summarize program data and check that you answered every question in the funder's guidelines. A person who knows the program should write the narrative and verify every number before you submit.

What is the cheapest way for a small nonprofit to start using AI?

Pick one recurring task, like the monthly newsletter, and build a written workflow around tools you already pay for. Many email platforms, CRMs and office suites now include AI features, so audit what you have before buying something new.

---

Cover photo: Radek Grzybowski, via Wikimedia Commons. Creative Commons CC0 1.0. Displayed cropped to fit.