The Giving Foundation · AI in Practice
What good looks like.
AI in your nonprofit,
done right.
Your AI Use Policy defines the guardrails. This resource answers the other question — the one everyone is actually asking: So what can I actually do with it?
🧭 The guiding principle: AI augments the human. Never the other way around.
This is not a policy document
It's a practical companion — written in plain language, organized around real nonprofit work. Everything here operates within the boundaries of your organization's Acceptable AI Use Policy. When in doubt, check there first, then ask your supervisor.
Area 1 of 3
Increase Operational Efficiency & Human Engagement
The administrative work that keeps your organization running — but doesn't need to consume your most valuable human hours. AI is a strong first-draft and organizing tool here.
Remember
AI is not always the answer. Before reaching for an AI tool, ask whether a simpler technology — a shared spreadsheet, a template, an existing workflow — would do the job just as well with less risk.
Meeting preparation & summaries
Use AI to draft agendas, compile discussion points from notes, and create structured meeting summaries for your records.
Try thisPaste your handwritten meeting notes into an AI tool and ask it to organize them into decisions made, action items, and open questions.
Job description drafts
AI can produce a solid first draft of a job description from a short bullet list of responsibilities — saving hours of starting from a blank page.
Try thisList 6–8 responsibilities and your desired qualifications. Ask AI to draft a job description in plain, welcoming language. You edit and own the final.
Knowledge management
Capture and organize institutional knowledge that currently lives in people's heads — processes, history, best practices — before it walks out the door.
Try thisInterview a long-tenured staff member about a key process. Paste the transcript and ask AI to turn it into a structured how-to document.
Volunteer coordination
Streamline volunteer communications, role descriptions, and scheduling materials so your team spends more time with volunteers and less time on logistics.
Try thisDescribe your upcoming volunteer event. Ask AI to draft a role breakdown, a welcome message, and a day-of checklist. Review and personalize before sending.
Research & landscape scanning
Use AI to quickly summarize publicly available information about trends, peer organizations, or policy environments relevant to your work.
Try thisAsk AI to summarize recent developments in a policy area your organization works in. Verify key claims before acting on them — AI can hallucinate.
Internal reports & board materials
Draft first versions of board reports, program updates, and internal summaries from your own data and notes — no confidential client information included.
Try thisShare your program metrics and key outcomes. Ask AI to draft a 1-page board update in plain language. Staff review and sign off before submission.
Area 2 of 3
Improve Human Decision Making
AI doesn't make decisions here — it helps humans make better ones. The distinction matters: AI surfaces patterns, options, and frameworks. Humans weigh the full picture and decide.
Human-Atop-The-Loop
In this area especially, the value of AI comes from augmenting your team's judgment — not replacing it. AI input should always be one voice among several, not the deciding one.
Program data analysis
Help your team make sense of program data — participation trends, outcome patterns, service gaps — so decisions are grounded in what's actually happening.
Try thisShare your anonymized program data (no client names or identifying info). Ask AI to identify 3 patterns worth discussing at your next team meeting.
Community feedback synthesis
Turn raw survey responses, focus group notes, or feedback forms into themes and insights that inform how you improve programs.
Try thisPaste anonymized survey responses. Ask AI to identify the top 5 themes and any notable tensions or contradictions in what people said.
Scenario planning support
When facing a significant decision, AI can help you think through scenarios, stress-test assumptions, and surface considerations you might have missed.
Try thisDescribe a strategic decision you're facing and what you know about the situation. Ask AI to outline 3 scenarios and the key risks in each. You and your team decide.
Policy & practice research
Quickly orient yourself in a new policy area — understanding the landscape, key players, and relevant debates — before diving deeper with primary sources.
Try thisAsk AI to explain a policy or funding area relevant to your work in plain language. Use it as orientation, not as your final source. Verify what matters.
Area 3 of 3
Enhance Human Service Delivery
AI can help your team show up better for the people you serve — by handling the logistical and administrative load so staff energy goes where it matters most. The human relationship remains central.
Important
The substance of any communication with clients, partners, or community members must remain human-generated. AI can help with logistics and preparation. It does not speak for your organization.
Translation assistance
Make materials accessible in community languages more quickly — flyers, intake documents, resource guides — with AI doing the first pass and humans reviewing.
Try thisPaste an English-language resource guide. Ask AI to translate it. Then have a fluent staff member or community partner review for accuracy and cultural nuance.
Emergency response preparation
Build communication frameworks, resource lists, and response protocols in advance so your team can move quickly when community need surges.
Try thisDescribe your most likely emergency scenario. Ask AI to draft a staff communication plan and a client notification template — both for human review and approval.
Program intake & scheduling docs
Streamline intake forms, screening checklists, and appointment scheduling materials so staff spend less time on paperwork and more time with clients.
Try thisDescribe your intake process step by step. Ask AI to redesign it as a cleaner checklist — then pilot it with staff before rolling out.
Manager preparation & coaching
Help managers prepare for conversations — performance check-ins, difficult feedback, 1:1 agendas — by thinking through their approach in advance. AI helps the manager think. The manager leads the conversation.
Try thisDescribe a situation you need to address with a staff member. Ask AI to help you think through key points and potential responses. You lead the meeting.
Know the boundaries
The line worth understanding.
Some of the most tempting AI use cases are also the ones your policy draws a firm line on. This isn't arbitrary — it's principled. The short version: AI doesn't speak for you. Your voice, your relationships, and your advocacy must remain authentically human.
✕
Communication with colleagues, partners, or clients
In any medium. Your words. Your relationship.
✕
Advocacy campaigns & external messaging
Your mission speaks in your voice, not AI's.
✕
Grant writing
Funders build relationships with your organization, not a language model.
✕
Addressing interpersonal or behavioral challenges
These conversations require human presence, judgment, and accountability.
✕
Financial management
Financial decisions carry fiduciary responsibility that cannot be delegated to AI.
✕
Any decision that bypasses Human-Atop-The-Loop
AI informs. Humans decide. Always.
A note on grant writing specifically
We know this one surprises people. The reason is principled: grant relationships are built on trust, specificity, and organizational voice — the exact things AI flattens. Funders fund you. That voice has to be yours.
A quick check
Three questions before you start.
Before using any AI tool for a task, run through these three questions. They take thirty seconds and they'll keep you on the right side of your policy — and your own judgment.
1
Is this tool on our approved list?
Using an unapproved AI tool — even a widely-used one — puts the organization at risk. If you're not sure what's on the approved list, ask your Technology Oversight contact before you start. If the tool you want to use isn't approved, ask for it to be evaluated rather than just using it anyway.
2
Does this task involve confidential information?
Never paste client names, case details, Social Security numbers, financial records, or anything that could identify a specific person into an AI tool. If you're not sure whether something counts as confidential, treat it as if it does. When in doubt, leave it out.
3
Will a human review and own this output?
AI output is a starting point, not a final product. Before anything AI-generated leaves your desk — a document, a summary, a translated resource — a human being must read it, verify it, and take responsibility for it. That's the Human-Atop-The-Loop principle in practice.
If you answered yes to all three
You're in good shape. Proceed thoughtfully, review the output carefully, and remember that AI is a tool — and you're the one accountable for how it's used.
"The most important thing we do is show up for people."
AI, used wisely, gives some of those hours back — and puts them where they belong.
Review the full AI Use Policy →