Free course: AI for nonprofit communications

Most nonprofit communications teams are one or two people doing the work of five. AI will not fix that, but it can take a real bite out of the work that eats your week: the third draft of a newsletter, the captions nobody has time for, the report that never becomes a story.

This course is about communications and marketing work specifically. It covers what AI is genuinely good at in comms, how to get output you can actually publish, which tools are worth knowing, what can go wrong, and a concrete plan to get started. It takes about 20 minutes and you do not need to buy anything to follow along.

*️⃣ Note

If you want the general foundations first, how AI actually works, the broad opportunities and risks across an organization, start with our free course on AI for nonprofit organizations. This course assumes you just want to get to work on comms, so it skips the general theory.

What can AI do for nonprofit communications and marketing?

The fastest way to find your own uses is to walk through your actual workflow and notice where AI fits. Here is the landscape, grouped by the work you already do. If you want to browse more widely afterwards, our database of the most common AI use cases in nonprofits covers every department.

AI for writing and drafting nonprofit content

This is where most teams start, and where the time savings are most immediate.

• First drafts of blog posts, op-eds, and impact stories from a brief and a set of facts you supply.
• Email subject lines and preheaders, generated as a batch of options to test rather than one guess. Email still drives a sizable share of nonprofit online revenue (2026 M+R Benchmarks), so this is high-leverage work.
• Newsletter copy built from your bullet points, then edited for voice.
• Personalized thank-you letters drafted from the gift amount, the fund, and a recent result.
• Press releases and media pitches from a fact sheet, tailored per outlet.
• Social media captions and platform variants, with several genuinely different angles rather than five versions of one idea.
• Ad and campaign copy, including search ads if you run Google Ad Grants.

AI for repurposing content into newsletters, social posts, and clips

This is the most underrated category, because the raw material is already sitting in your drive.

• Turn one long piece into many: an annual report or webinar becomes a newsletter section, several social posts, and a donor email.
• Turn a program report into a story your supporters will actually read.
• Transcripts and captions from any recording, which is both a time-saver and an accessibility requirement.
• Meeting notes into decisions and action items.
• Plain-language rewrites of dense pages, to a target reading level.
• Long reports into short audio overviews your board can listen to on a commute.

AI for research: funders, journalists, and competitor messaging

Useful, but this is also the category where AI most often invents things, so verification is not optional.

• Funder and sector research, using a tool that cites its sources so you can check them.
• Journalist and media-list research: who covers your cause, and what they published recently.
• Competitor and peer messaging analysis: how similar organizations describe their work, and which positioning is unclaimed.
• Audience and persona research to sharpen who you are writing for.
• Survey design, drafting unbiased questions before you send anything.
• SEO keyword and content-gap audits to find what people actually search for.

AI for analyzing surveys, comments, and analytics

• Open-text survey synthesis into themes and representative quotes, which is otherwise a day of manual reading.
• Social listening and comment synthesis into themes and sentiment.
• A plain-language monthly summary of your analytics, instead of a dashboard nobody opens.
• A/B test read-outs that tell you what to try next.

AI for accessibility: captions, alt text, and translation

These deserve their own mention because they are usually the first things cut for lack of time.

• Alt text for every image you publish.
• Captions and transcripts for every video.
• Translations of key content into the languages your community actually speaks.
• Reading-level rewrites so your materials work for more people.

New projects AI makes possible for small nonprofit teams

Everything above is work you already do. The bigger opportunity is work you currently skip because it was never affordable.

• Translate everything, not just the flagship page. A newsletter in three languages used to mean a translation budget.
• Thank every donor personally, not only major gifts.
• Turn every webinar into a full content pack instead of letting recordings die in a folder.
• Publish far more impact stories, because the bottleneck was drafting time, not material.
• Answer the same 20 questions automatically on your site, freeing your inbox for the messages that need a human.

⬆️ Important

The best first AI project is usually boring: repetitive, low-stakes, based on material you already have permission to share. Save the ambitious public-facing ideas until you have some experience and the guardrails in place.

AI agents for nonprofit communications

Everything above treats AI as an assistant: you ask, it answers, you decide what to keep. Agentic AI goes a step further. You give an agent a goal, and it plans and carries out the steps itself across your files, your apps, and the web, pausing for your approval on the actions that matter. Instead of prompting turn by turn, you delegate a whole task and review the result.

That opens work that used to be too slow or fiddly to attempt: multi-step jobs that meant a dozen separate chats, constant copy-pasting between tools, or an afternoon of manual clicking. Part of what makes agents more capable is that they can do more than write text back to you: they can connect to your other tools (through connectors, the emerging MCP standard, or APIs), create and edit real files, and run saved routines called skills. The most capable general-purpose options right now are knowledge-work agents like Claude Cowork or ChatGPT Work. Most major AI vendors are shipping one, so check what your current tools already include before paying for a new seat.

This is a fast-moving space, so treat the list below as a few examples, not the full set. Some of what agents make newly practical for a communications team:

• Connect an agent to the tools you already use. Through connectors, MCP, or APIs, an agent can work across your CRM, inbox, calendar, analytics, and scheduler in one task, for example reading last week’s email stats and drafting the performance summary.
• Create and edit real files, not just chat replies. An agent can produce and revise documents, spreadsheets, and images directly in your drive, so you get a finished draft file rather than text to copy out of a chat window.
• Save a recurring job as a reusable skill. Teach an agent your repurposing recipe or your brand-voice checklist once, save it as a skill, and it runs that job the same way every time.
• Turn one webinar into a full content pack. Hand over the recording and the slides and get back a draft blog post, a clip list, social captions, and a newsletter section in a single task, instead of running each through its own chat.
• Complete research briefs in minutes. A deep-research mode (in ChatGPT, Claude, or Perplexity) runs a multi-step scan of funders, journalists, or peer messaging and returns one cited summary. This is the easiest first taste of an agent, and it may already be in a tool you have.
• Get a standing brief on a schedule. An agent gathers last week’s mentions, sector news, and your analytics into one short summary automatically, with no prompting each time.
• Hand off routine browser and admin work. Refresh your organization’s listing across directories, update event or volunteer postings on several sites, or pull data from a platform with no export button, using a browser agent such as Claude for Chrome or ChatGPT Work’s built-in browser.

⚠️ Warning

Because an agent acts on its own, it carries risks a chatbot does not. The main ones to know, among others:

• It acts, so mistakes ship. A wrong edit, post, or send can go out with no person in the loop. Keep an approval gate on anything that publishes, emails, pays, or deletes, and start any agent in read-only mode.
• Hidden instructions can hijack it (prompt injection). A web page or document the agent reads can contain instructions it may follow without you realizing. Give it only the access a task needs, nothing more.
• Connected tools and add-ons widen what can go wrong. Every tool, connector, or skill you grant an agent is more it can reach and more code you are trusting, so connect only tools you need and install skills or plugins only from sources you trust.
• Wider access means wider data exposure. An agent reaching into your drive, inbox, or CRM can touch far more sensitive data than a single pasted prompt, so your data rule matters more here, not less.

Our fuller guide to AI agents for nonprofits covers what they can do today and the guardrails to set, and agent skills, plugins, and connectors explains how to extend one safely.

How to write better AI prompts for nonprofit communications

This is the part most people skip, and it is the difference between “AI writes generic junk” and “AI saves me half a day”. These habits matter far more than which tool you pick.

Build a reusable context pack

The single biggest quality jump comes from telling the AI who you are, who you are writing for, and what you are trying to achieve. Write this once and reuse it forever. A good context pack includes:

• Your mission and what you actually do, in plain language.
• Your audience, and what they care about.
• Your voice, described concretely (warm, plain, no jargon, no guilt appeals).
• Things you never say, including terms your community rejects.
• Two or three examples of your best past work.

Prompting techniques that actually improve output

• Show, do not describe. Pasting two or three of your strongest past newsletters teaches the AI your voice better than any adjective. If you only do one thing from this course, do this.
• Ask for options, not an answer. “Give me eight subject lines across four different angles” beats “write a subject line” every time. You are better at picking than the AI is, so let it generate and you select.
• Say what you do not want. Constraints improve output sharply: no statistics that are not in the source, no exclamation points, under 120 words, no invented quotes.
• Give it a role and an audience. “You are an editor at a small environmental nonprofit writing for monthly donors over 60” produces noticeably different work than an unframed request.
• Work in a conversation, not one perfect prompt. Get a draft, then steer it: shorter, warmer, lead with the volunteer, cut the third paragraph. Iterating is faster than engineering the perfect instruction.
• Make it critique itself. Ask the AI to review its own draft against a checklist (is the ask clear, is the opening specific, would a first-time reader understand this) and then rewrite. This catches a surprising amount.
• Handle long jobs one piece at a time. Ask for one item, give feedback, then continue. Quality drops when you ask for twelve things at once.
• Save what works. A prompt that produced a good newsletter is an asset. Keep a document of working prompts and your team stops reinventing them.

If you want to go further on this, our guide to prompt and context engineering covers the techniques above in much more depth.

✅ Example

A reusable prompt skeleton you can adapt for almost any drafting task:

# CONTEXT
> Our organization: [NAME], working on [CAUSE].
> Our voice: [e.g. warm, plain-language, hopeful. No jargon, no guilt appeals].
> Audience: [e.g. monthly donors, mostly over 50, who care about local impact].
> Examples of our style: [paste 2 short excerpts of your best past work].

# SOURCE MATERIAL
> [paste the approved facts, report, or notes. Nothing confidential.]

# REQUEST
[e.g. Draft a 200-word newsletter section and 3 social captions with
different angles.]

# CONSTRAINTS
Use only facts from the source material. Do not invent numbers, names,
quotes, or claims. Flag anything you were unsure about at the end.

The best AI tools for nonprofit communications (and how to get them cheaper)

You do not need to buy anything to start. A free general assistant covers most of what is above. Every tool below is an example within its category, not a recommendation to adopt: the field moves fast, and the practice matters more than the vendor.

• General assistants for drafting, summarizing, and rewriting: ChatGPT, Claude, or Gemini. Start here.
• Research with citations: Perplexity, or the search and deep-research modes inside the general assistants.
• Transcription and captions: Otter, Descript, or Fireflies.
• Design and graphics: Canva or Adobe Express.
• Short video clips from long recordings: Descript or Opus Clip.
• Readability and editing: Hemingway Editor or Grammarly.
• Email marketing with built-in AI: Mailchimp or Brevo.
• Social scheduling with AI assistance: Buffer or Hootsuite.
• Connecting tools together: Zapier, Make, or n8n, for example so a new blog post automatically drafts social posts into an approval queue.

For a fuller, regularly reviewed list, see our AI tools for communications and marketing guide, or the broader best AI tools for nonprofit organizations.

*️⃣ Note

Before you pay for anything, check the nonprofit programs. TechSoup offers discounted and donated software to eligible organizations, and Google for Nonprofits includes Google Workspace and the Ad Grants program for search advertising. Many AI tools also have nonprofit pricing that is not advertised on their pricing page, so it is worth asking.

AI risks for nonprofits: what can go wrong and how to prevent it

None of this is a reason to avoid AI. It is a reason to use it with a few habits in place. Our database of key AI risks for nonprofits covers each of these in more detail, with mitigation strategies.

• Invented facts and fake sources. AI states false things with complete confidence, including statistics, quotes, and citations that do not exist (Charity Digital on AI hallucinations). Verify everything factual against a primary source before publishing.
• Data leaving your control. Many consumer tools may use what you type to improve their models unless you change the settings. Never paste donor personal data, identifying details about the people you serve, unreleased announcements, or anything under NDA.
• AI images of the people you serve. Photorealistic AI images of beneficiaries raise consent and dignity problems, and research on charity imagery found that labelling the images did not prevent backlash, because the objection is to the depiction itself (University of East Anglia).
• Bias and stereotyping. Image tools tend to over-represent stereotypes, sometimes more extremely than reality (Washington Post analysis). Review anything depicting people or communities with extra care.
• Voice clones and likeness. Cloning a real person’s voice or face without written consent is a reputational and legal risk, not a shortcut.
• Losing your voice. Unedited AI copy drifts toward generic. If everything you publish starts sounding the same, that is the drift, and the fix is heavier editing and better examples in your context pack.
• Disclosure and donor trust. A supporter who discovers an undisclosed AI-written “personal” appeal may not give again. Disclose where your audience would feel misled to learn it later.
• Automation without a human. Any automation that sends should save a draft instead. You keep almost all the time saved and none of the risk of an unreviewed message going out in your name.
• Accessibility errors. Auto-captions and alt text still need a human pass, especially for names, places, and anything quotable.

⚠️ Warning

One of the most useful things you can do early is write a short no-go list: the handful of things your organization decides in advance never to do with AI. Deciding now removes the pressure to cut a corner later, when a deadline is looming. A reasonable starting list for a comms team:

• Never publish a fabricated beneficiary story, quote, or testimonial.
• Never use photorealistic AI images of identifiable people you serve.
• Never run a chatbot that hides that it is AI or cannot hand off to a person.
• Never publish an AI-generated statistic or fact without checking a primary source.
• Never auto-send an AI-drafted message without a human reading it.

Next steps: how to start using AI in your comms team

Reading about AI changes nothing on its own. Here is a realistic order of work, from this week to the next few months.

1. Pick one repetitive, low-stakes task and try it this week. Use material you already have permission to share. Repurposing an existing update into a newsletter section and a few social posts is the usual best starting point.
2. Build your context pack (mission, audience, voice, things you never say, two examples of your best work). Save it somewhere shared. Every future prompt gets better from this one document.
3. Write your data rule: the short list of what nobody pastes into a consumer AI tool. Donor personal data, identifying details about the people you serve, and anything unreleased or under NDA are the usual entries.
4. Agree a short no-go list with whoever owns communications decisions, using the list above as a starting point.
5. Name who reviews AI output before anything is published. This one decision is what turns “we should check it” into something that actually happens.
6. Run a two-week pilot on that first task. Write down what “working” means before you start (usable drafts most of the time, less time than doing it by hand) and when you would stop (for example, it invents facts twice, or fixing output takes longer than writing from scratch). Then keep it, adjust it, or drop it honestly. Our checklist for new AI pilots and projects walks through what to decide before you start.
7. In your first month, expand to a handful of uses rather than everything at once, and keep a shared document of the prompts that worked. Most of the compounding value comes from reuse, not from new tools.
8. In months two and three, write a one-page AI policy covering your data rule, your no-go list, who reviews what, and when you tell your audience that AI was involved. One page that people read beats ten pages that nobody opens. Our AI policy template gives you a ready-to-adapt starting point.
9. Bring your colleagues in. Show two or three concrete before-and-after examples from your own pilot. That persuades far better than an article about AI, and it surfaces the objections you need to answer.
10. Set a review date every few months. Tools, prices, and capabilities change quickly, so a use that failed may be worth revisiting, and your policy will need updating.
11. Go deeper when you are ready to do this properly. Our full AI course for nonprofit communications covers the complete use catalog, tool selection and budgeting, the full risk and disclosure framework, team roles and buy-in, and how to measure whether a pilot actually worked. We also have equivalent courses for fundraising, operations, programs, HR, and leadership.
12. Get help if you would rather not do it alone. We offer AI consulting for nonprofits, including a free first consultation to talk through your situation and where AI would genuinely help.

Whatever you do next, the first step matters more than the plan. Pick one task, try it, and decide with evidence rather than opinion.