AI for Nonprofits: A Practical Guide
AI for nonprofits without the hype: the real ChatGPT, Claude, Gemini and Copilot nonprofit discounts, safe uses for grants and donors, and a one-page AI policy.
Key takeaways
- AI for nonprofits is mostly a drafting, summarizing and admin tool: grant narratives, donor letters, board minutes and reports, always with a human reviewing before anything goes out.
- The big four platforms all have nonprofit pricing in 2026: Gemini and NotebookLM at no cost in Google Workspace for Nonprofits, up to 75% off Claude and ChatGPT business plans, and a discounted Microsoft 365 Copilot add-on.
- Nearly every nonprofit already uses AI, but almost half have no policy. A one-page policy plus one approved platform beats a dozen personal accounts holding donor data.
AI for nonprofits means using generative AI and simple automation to draft grants, donor communications and reports, summarize meetings and data, and cut the admin work that eats staff time. The sensible approach for a small or mid-size nonprofit is to claim one discounted business-grade platform, write a one-page policy, keep donor and client data out of personal accounts, and have a person review every output.
I run technology for businesses from Houston. My managed IT company, TMG, has a nonprofit practice, and I founded Talos Automation, which builds AI workflow automation, so I have a commercial interest in organizations adopting this and you should weigh what follows accordingly. What I can offer is the operator view: what the nonprofit pricing programs actually include, where AI saves real hours, and where it quietly creates risk for an organization whose most valuable asset is donor trust.
Most of what ranks for this topic is written by fundraising software vendors, and each guide ends with a demo button for their CRM. This one does not recommend a product. It gives you the real 2026 discount programs with their sources, a short list of uses worth starting with, the grant-writing rules funders are now setting, and a policy you can adapt this week. If you want the general version for any small organization first, my operator's guide to AI automation strategy covers the same thinking.
How Are Nonprofits Using AI Today?
Almost all of them use it, and very few use it well. In a survey of 346 nonprofits fielded in December 2025, Virtuous and Fundraising.AI found that 92% use AI tools in some capacity, 81% use them individually and ad hoc, only 4% have documented, repeatable workflows, and just 7% report major improvements in their ability to achieve their mission.
That gap is the whole story. AI in most nonprofits today is a development director pasting a donor list into a personal ChatGPT account at 10 p.m. to get a draft appeal out. It saves an hour, it is invisible to the executive director, and nobody has asked where the donor list went. The 2026 Nonprofit AI Adoption Report also found that 47% of organizations have no AI governance policy at all. I have written about this pattern in businesses as shadow AI, and nonprofits have a sharper version of it because so much of the work is done by volunteers on their own devices.
The organizations that get real value do three boring things: they pick one platform, they write down what may and may not go into it, and they turn the two or three tasks they repeat every week into a documented workflow instead of a personal habit.
What Are the Best AI Tools for Nonprofits?
For most nonprofits the best AI tool is the one bundled with the productivity suite you already run, bought through its nonprofit program. If you live in Google Workspace, start with Gemini and NotebookLM. If you live in Microsoft 365, start with Copilot. Add a standalone assistant like ChatGPT or Claude only when a specific workload needs it.
Every major vendor now has a nonprofit offer, and the differences matter more than the marketing suggests. Here is what each program says, from the vendor's own pages, as of November 2026.
The conclusion from that table: start with the no-cost or bundled option inside the suite you already pay for, then buy discounted seats of one standalone assistant for the two or three people who write the most. Paying full consumer prices on personal cards is the most expensive and least safe way a nonprofit can use AI.
- Google: Google's nonprofit help center lists the AI features included in Google Workspace for Nonprofits, including the Gemini app, Deep Research, Gems and NotebookLM, with enterprise data protections. Note what is not included at no cost: Help me write in Gmail and Docs and the Gemini side panel need a paid edition.
- Anthropic: the Claude for Nonprofits announcement of December 2, 2025 offers up to 75% off Team and Enterprise, connectors for Blackbaud, Candid and Benevity, and a free AI Fluency for Nonprofits course built with GivingTuesday.
- OpenAI: the OpenAI for Nonprofits page was updated on February 6, 2026 to offer up to 75% off ChatGPT Business or Enterprise; you request it through OpenAI's sales team.
- Microsoft: the Microsoft nonprofit offers guide lists Microsoft 365 Copilot at $25.50 per user per month on annual billing. It is an add-on, so each user also needs a qualifying Microsoft 365 license.
If you are still choosing a suite, pick the one your staff already know; switching platforms to chase an AI feature rarely pays back for a small team. For the standalone assistants, my Claude vs. ChatGPT comparison covers the trade-offs, and the SMB tech stack guide shows where AI sits in the wider setup.
Is ChatGPT Free for Nonprofits?
No. ChatGPT has a free consumer tier anyone can use, but OpenAI does not give nonprofits a free business plan. What it offers is a discount of up to 75% on ChatGPT Business or Enterprise. The free tier runs on a personal account the organization does not control, which is the wrong place for donor or client information.
The practical difference between a personal account and a business plan is not the model, it is the controls. A business or nonprofit workspace gives you an admin who can add and remove people, single sign-on, settings that keep your inputs out of model training by default, and the ability to take a departing volunteer's access away on their last day. That last one is the control nonprofits most often discover they needed after the fact. My AI tools stack for SMBs covers how to choose seats without overbuying.
How Can Nonprofits Use AI for Grant Writing?
Use AI for grant writing as a drafting and editing assistant, never as the author. It is good at turning your program notes into a first-draft narrative, tightening a 900-word answer to a 500-word limit, building a logic model outline, and checking a draft against the funder's stated priorities. Your program facts, outcomes data and voice have to come from you.
Funders are starting to write rules about this, and the federal ones are the strictest. In notice NOT-OD-25-132, the National Institutes of Health said it will not consider applications substantially developed by AI, or with sections substantially developed by AI, to be the original ideas of the applicants, that it uses detection technology, and that findings after award can lead to disallowed costs, withheld awards or termination. The same notice caps each principal investigator at six applications per calendar year from the September 25, 2025 receipt date. Most small nonprofits apply to foundations rather than NIH, but private funders read the same headlines.
- Read the funder's AI rules first. Check the request for proposals, the funder's website and any certification you sign. When in doubt, ask the program officer.
- Feed it your own material. Give the assistant your past successful proposals, program data and outcome reports, not a blank prompt, so the draft reflects your actual work.
- Draft, then rewrite in your voice. Treat the output as a rough first draft. A grant that reads like everyone else's AI draft is a grant that scores lower.
- Verify every number and citation. AI tools invent statistics and sources with total confidence. Every figure in a proposal should trace to your data or a named source.
- Keep a note of how you used it. A two-line record of which sections used AI assistance makes any later funder question easy to answer honestly.
Followed in that order, AI turns grant season from a series of all-nighters into a review job, without putting your organization's credibility with a funder at risk.
Where Does AI Save Nonprofit Staff the Most Time?
The biggest savings are in recurring writing and summarizing work that follows a pattern: donor acknowledgments, newsletters, board packets, meeting minutes, volunteer communications and impact reports. These tasks repeat every week or month, have a template, and are reviewed by a person anyway, which makes them low-risk places to start.
- Donor stewardship: personalized thank-you letters from a gift record and a template, with a staff member signing off on each one. Keep gift amounts and donor notes inside your approved platform.
- Board and committee work: summarizing meeting transcripts into minutes and action items, and drafting the board packet cover memo from the month's reports.
- Impact reporting: turning program spreadsheets into a plain-language summary and a first draft of the annual report narrative.
- Volunteer coordination: drafting shift reminders, onboarding emails and FAQ answers so the coordinator is not rewriting the same message every week.
- Phones and inquiries: an AI receptionist can answer common questions about hours, donations drop-off and services after hours and hand anything sensitive to a person. My guide to AI receptionists for small business covers how to evaluate one.
Pick two of these, not five. My rule for any organization is the one in what your business should automate first: start where the volume is high, the risk is low and someone already complains about the work. Once a task is stable as a prompt, it can become an actual workflow that runs on a trigger, which is what my agency's done-for-you workflow automation builds, with the standing disclosure that it is my company.
What Are the Risks of AI for Nonprofits?
The main risks are leaking donor or client data into tools the organization does not control, publishing invented facts in grants and reports, and losing the human voice donors respond to. Nonprofits carry extra exposure because they often serve vulnerable people and depend entirely on public trust, so a single privacy incident costs more than the hours AI saved.
- Data leakage: client case notes, health details, immigration status and donor wealth data must never go into a personal account. Organizations serving clients in health programs may also have HIPAA obligations; the HHS HIPAA guidance for professionals is the place to confirm whether that applies to you.
- Hallucinated facts: invented statistics in a grant or impact report damage credibility with funders far more than a late submission.
- Voice and trust: donors give to people. Appeals that read as generic machine copy underperform, so AI should draft and a person should finish.
- Volunteer access: volunteers rotate. If they use AI on personal accounts with your data, that data leaves when they do.
- Security basics: an AI assistant can only see what your file permissions allow, so messy shared drives become messy AI answers. Fix permissions and turn on multi-factor sign-in before you add AI to the suite; my SMB cybersecurity checklist covers the minimum.
Does a Nonprofit Need an AI Policy?
Yes, and it can fit on one page. A nonprofit AI policy names the approved tools, lists the data that must never go into any AI tool, requires a person to review anything published or sent to a funder, and says who to ask. The board should approve it, because AI use touches donor trust and the board owns that.
Here is the outline I use. It is a shortened version of the template in my guide to writing an AI policy for a small business, adjusted for nonprofits:
- Approved tools. Name the one or two platforms on the organization's nonprofit plan. Everything else is not approved for organizational data.
- Restricted data. Client records, health information, donor giving history and contact details, staff HR records and anything under a confidentiality agreement stay out of AI tools unless the tool is approved for that data in writing.
- Human review. Every grant, appeal, report, public post and client-facing message is reviewed and owned by a named staff member before it goes out.
- Funder and donor transparency. Follow each funder's AI rules, and be honest if a donor or funder asks how you use AI.
- Volunteers. Volunteers use organizational accounts for organizational data, and access ends on their last day.
- Owner and review date. One person owns the policy, and it is reviewed every six months because the tools change that fast.
If you want the governance layer above the policy, such as risk tiers and how to review new tools, my AI governance framework for small business scales down to a nonprofit board without much editing.
How Should a Small Nonprofit Get Started With AI?
Start in 30 days with four steps: claim the nonprofit offer on your existing productivity suite, move anyone using personal AI accounts for work onto it, approve a one-page policy, and pilot two workflows with a clear measure of hours saved. You do not need a consultant, a data strategy or a new CRM to do any of that.
- Week 1: verify your organization for Google for Nonprofits or Microsoft's nonprofit program if you have not already, and claim the included AI features.
- Week 2: ask staff and key volunteers which AI tools they already use for work. Expect the answer to be more than you think, and treat it as information, not a disciplinary matter.
- Week 3: take the one-page policy to the board or executive committee, and set up the approved platform with single sign-on and multi-factor sign-in.
- Week 4: pilot two workflows from the time-saving list above, time them for a month, and keep, change or drop them based on the numbers.
If you want a quick self-check before week 1, my 12-point AI readiness checklist works the same for a nonprofit as for a business. In Houston, the groundwork underneath all of this, such as identity, backups and device security, is what my team at TMG covers in its IT services for Houston nonprofits.
The Bottom Line
AI for nonprofits is already happening in almost every organization, mostly on personal accounts and without a policy. The fix is cheap: every major platform now has nonprofit pricing, Google includes Gemini and NotebookLM at no cost, and a one-page policy takes an afternoon. Start with the suite you already run, move staff onto one approved platform, use AI for drafting and summarizing with a person reviewing every output, follow each funder's rules on grant writing, and keep donor and client data out of anything you have not vetted. Do that, and AI gives your staff back hours for the mission instead of creating a privacy problem you have to explain to donors.
If you want a second opinion on your organization's setup, book a free strategy call and bring the one recurring task your team complains about most.
Frequently Asked Questions
How can nonprofits use AI?
Most nonprofits get the fastest return from first drafts of grant narratives and reports, donor thank-you letters and newsletters, meeting and board-minute summaries, volunteer scheduling messages, and pulling figures out of program data. A person reviews every output, and nothing confidential about donors or clients goes into a tool the organization has not approved.
Is there a ChatGPT discount for nonprofits?
Yes. OpenAI runs an OpenAI for Nonprofits program, and a February 2026 update to its announcement says eligible nonprofits can get up to 75% off ChatGPT Business or ChatGPT Enterprise through its sales team. The exact rate depends on verification and seat count, so get a written quote before you budget.
What are the best free AI tools for nonprofits?
Google Workspace for Nonprofits includes the Gemini app and NotebookLM at no cost for up to 2,000 users with enterprise data protections. Microsoft 365 work accounts include Copilot Chat. Those two cover drafting, summarizing and research for most small organizations before you pay for anything else.
Can nonprofits use AI to write grant proposals?
Yes, as a drafting and editing assistant, but not as the author. Funders increasingly ask about AI use, and the National Institutes of Health says it will not treat applications substantially developed by AI as the applicant's original ideas. Use AI for outlines, boilerplate and edits, keep your own program facts and voice, and check each funder's rules.
Does a nonprofit need an AI policy?
Yes, even a one-page one. A December 2025 survey of 346 nonprofits by Virtuous and Fundraising.AI found 92% already use AI but 47% have no AI governance policy. A short policy naming approved tools, banned data and the review rule protects donors, clients and the organization's reputation.
Is donor data safe in AI tools?
Only in tools on a business or nonprofit plan whose terms say your inputs are not used to train models, with retention controls and single sign-on. Free consumer accounts on personal email are where donor lists and client case notes leak. Treat donor and client records as restricted data that never goes into an unapproved tool.