How nonprofits can use AI to give their team time back

What would you do with 4 to 8 extra free hours every week?

I want to start with a question, and I'd like you to answer it before you read on.

If everyone on your team got four hours back next week, what would you do with them? What about six? Eight?

I don't mean catching up on email. I mean the things that keep sliding to next month: calling the donor who gave for the first time, finally reading the program feedback forms, planning the volunteer training you keep meaning to improve, sitting down for an hour with a new staff member or volunteer to show them how things really work, so they feel welcome and want to stay. The details you know would make a difference, if anyone had the time to pay attention to them.

Hold on to your answer. AI might just be the key to getting that time back.

Where things stand

Almost every nonprofit is already using AI in some way. In a survey of 346 nonprofits taken in December 2025, 92% said so. Only 7% said it had made a major difference to what they can achieve. (Virtuous, 2026 Nonprofit AI Adoption Report)

The same report shows where the gap comes from: 81% use AI ad hoc, one person at a time, with no shared way of working. People are a little faster on their own tasks. The organization hasn't changed.

That is a normal stage, not a mistake. It is how every tool we now take for granted arrived.

A long line of tools

Think about what has sat on nonprofit desks over the years. The typewriter made a letter something you could produce in an afternoon instead of a day. The computer let you fix a typo without starting the page again. The spreadsheet turned a week of adding up a budget by hand into a formula. Software for donors, volunteers and programs meant the history of a relationship lived somewhere other than one person's memory.

None of those tools served a single person in your community. Each one took a piece of repetitive work off someone's hands, and the time went back to people.

AI is the next one on that list. It is very good at the tedious, repetitive and first-draft parts of the job. It is not in charge of anything. A person decides what it works on, checks what it produces, and owns what goes out.

What it looks like in practice

Here is how I use it every week.

• A check before anything goes out. Before an email, a page or a plan leaves my hands, I ask AI to read it the way the recipient will, and to tell me what's unclear, wrong or missing.
• A morning digest. Instead of opening six sites, I start the day with one short summary of the news in technology, AI and nonprofits, and what's worth my attention.
• Email triage, with drafts. AI sorts what needs a reply from what doesn't, and writes a first draft of the replies. I review, edit and send every one myself.
• Trends in a campaign. It's connected to the tools that track my campaigns, so I start with its first read of what changed. Then I ask questions and dig deeper until I understand what's working and what isn't.
• Talking a decision through. When I'm weighing whether to put my time into one campaign or channel over another, I lay out both and ask it to argue the pros and cons of each. It doesn't decide for me. It helps me see the choice more clearly.

None of these is impressive on its own. Together they give me back 10 to 15 hours every week.

The real gains come when a whole team shares what works. I wrote instructions that check a draft against our brand voice: the words we use and avoid, how warm or plain we sound, and the claims we can back up. Before anything goes out, it flags what doesn't sound like us. Put those instructions in a shared document, and every volunteer, intern and staff member gets the same check. Nobody has to rewrite everyone else's work.

How to grow into it: name an AI champion

You don't need a strategy document or a new system. You need one person with a little protected time.

1. Name a champion. Choose someone curious and trusted by the team. It doesn't have to be the most technical person, or you. Their job is to learn the tools and bring back what works.
2. Protect their time. Two to three hours a week, on the calendar, treated like any other commitment. This is the step that usually gets skipped, and it is the one that makes the difference. Learning a tool takes time before it gives time back.
3. Give them a plan. Start with two things:
   • A list of the most repetitive work. Ask each person on the team to write down the three tasks they do most often that feel tedious: the weekly report, the meeting notes, the same answer to the same question. The champion collects them and picks one to start with.
   • A short script for a staff meeting. Twenty to thirty minutes once a week, depending on the size of your team: here is one task we tried with AI, here is how it went, here are the instructions we gave it, if you want to try them. Then one question for the room: what should we try next?

Keep a person in charge. Agree early on what never goes into an AI tool (donor names and gifts, client details, anything private) and who checks the result before it leaves the building. If you want a starting point for the rules, Ember has a simple AI use policy you can copy.

More time for the work that matters

The goal was never more AI. It was more hours for the work I enjoy and the work with the greater impact. Finding ways to get there feels like joy.

Kindling for this week: one small step to try

The goal

Name your AI champion, put two hours on their calendar for next week, and have each person on the team list the three most repetitive tasks they do.

The metric

Is the champion's time booked, and how many people turned in their list?

How you'll know it worked

The time is on the calendar, at least half the team sent their three tasks, and the champion has picked one to try first.

If it doesn't work out

If the lists didn't come in, ask at your next staff meeting instead and write the answers down together in five minutes. If no one wants to be champion yet, take the first two weeks yourself and hand it on.

Tried it? Email Brian at hello@donewithember.org and tell him how it went: what worked, what didn't, and what you'd like the next entry to cover.