TL;DR
- AI adoption stalls when staff feel their belonging is threatened. People can’t be curious and afraid at the same time.
- Before adding any tools, run one diagnostic: is your tech ahead of your culture, or is your culture ahead of your tech?
- Four practical steps build readiness: address the belonging question, make complaining safe, find your first chair, and hold short regular conversations.
- AI will change jobs. The fundraisers who lean in early will be the ones who adapt.
Most nonprofit teams are running two conversations about AI at once.
- The loud one covers tools, prompts, and whatever feature launched this week.
- The quiet one happens in your head during a staff meeting: Am I training my own replacement?
Until leaders deal with the quiet conversation, the loud one goes nowhere. That’s the core of a recent episode of The Responsive Lab, where hosts Scott Holthaus and Carly Berna sat down with Tim Lockie, founder of The Human Stack, who’s spent his career helping nonprofits get more impact out of their people, data, and systems. The conversation covered why AI rollouts keep stalling, how to tell whether your team is actually ready, and what to do if it isn’t.
Why Most AI Rollouts Fall Apart Before They Start
The standard explanation is change management. Plenty of leaders have never walked a team through a technology shift, so they avoid the discomfort, and AI never reaches the actual fundraising work. That explanation is true. It’s also incomplete.
Underneath the change management gap sits an identity question that almost nobody addresses out loud.
“Behavior stems from what we believe about how we belong,” said Tim Lockie, founder of The Human Stack. “Belonging is the base code of that. That’s the deepest human need. And AI, corporately and in our jobs, is actually putting a threat risk on our belonging.”
When people are wondering whether they’ll still have a seat at the organization in two years, no amount of training fixes the resistance. As Lockie put it, “It is very hard to be curious and afraid at the same time.”
Most consultants and AI policies skip this layer entirely. They cover prompts, governance, and tool selection while the deeper question sits unanswered. If your team seems checked out on AI, start by assuming the problem is fear, and address the fear before the software.
A Simple Diagnostic Before You Do Anything Else
One question tells you most of what you need to know: is your tech in front of your culture, or is your culture in front of your tech?
If your tech is out in front, every new tool makes things worse. Systems that were supposed to save time create disruption instead. Staff quietly stop trusting the official database and build side spreadsheets, so you end up with eight informal CRMs instead of one. Sound familiar?
The same logic applies to your data. If your team doesn’t trust the data in your CRM, you shouldn’t trust what AI does with that data either. Layering AI on top of a culture that’s already behind its tools widens the gap.
So before you buy anything new, get honest about which side of that line you’re on. If culture is behind, the next four steps close the distance. None of them requires a big budget, and the cultural shift moves faster than most leaders expect.
4 Steps to Build a Culture That Can Actually Use AI
Step 1: Address the Belonging Question Directly
Say the uncomfortable part in a team meeting before anyone has to ask it.
Yes, AI will take some jobs. Pretending otherwise is disingenuous, and your team can tell. Offer a path instead: people who learn these tools for the work they already know how to do are positioning themselves well, no matter what comes next.
History backs this up. When cars arrived, the saddle makers who thrived became seat makers. Same craft, same leather skills, new context. And notably, they were the ones who actually rode in cars, noticed the seats were uncomfortable, and realized they could do better. The ones who stayed angry on the sidelines got left there.
Give your team explicit permission to ride in the car.
Step 2: Make It Safe to Complain
Complaining is a sign of hope. People who have given up stay silent.
Keep this separate from your help desk queue. The goal is a low-friction way for staff to say “this isn’t working for me” or “I don’t know how to do this thing” without it turning into a performance conversation. The smaller the frustrations people feel safe voicing, the faster you can fix them, and the more your team learns that raising a hand is rewarded rather than punished.
That psychological safety is the soil curiosity grows in. Skip it and every AI tool you roll out lands on a team that’s privately working around it.
Step 3: Find Your AI First Chair
Most departments have someone who’s already experimenting and somehow always knows the answer when a tool breaks. Organizations have called these people accidental admins. Tim calls them the first chair, borrowing from the orchestra: the musician who leads a section and helps everyone around them play better.
The first chair is already leading. The leader’s real job is to follow them, publicly. Recognize what they’re doing, tell the team this is the direction, and write it into their job description. Then meet with them weekly and ask two questions: What else does our team need? Who on our team isn’t getting it yet?
This pairing, a leader who clears the way and a first chair doing the tactical work on the ground, is where tech adoption succeeds or fails. Teams that get those two roles aligned tend to succeed with AI for the same reason they succeeded with every system before it.
Step 4: Hold Short, Regular Conversations
Twenty minutes, twice a week. The agenda is one question: what’s not working?
These small conversations do double duty. They surface problems while the problems are still cheap to fix, and they reveal who your first chairs are, because someone in the room always knows the answer. They also build the trust that makes shared data usable. Most of what gets labeled a data silo is really just uncoordinated data, teams playing the same notes at different times with no one conducting. Coordination gets built in exactly these kinds of small, repeated conversations, the way an orchestra builds confidence by practicing together rather than by hiring one great soloist.
Consistency matters more than length here. A standing twenty-minute rhythm beats a quarterly retreat every time.
The Honest Answer on Job Loss
It would be easier to end with reassurance that nothing will change. It would also be wrong, and your staff would sense it.
Some jobs will change shape, and some will disappear. Nobody can predict which, for whom, on what timeline. What history does show is that humans have run this exact panic loop with every major technology, from books to radio to computers (computer used to be a job title held by a person). The fear is always the same, and so is the outcome: people adapt, usually by engaging early rather than waiting.
And engagement starts with your own fear. Accept that this shift is happening whether or not anyone wants it, then decide what it means for your work. Learning to use AI in the work you already know is almost never the wrong move, whatever happens next.
Where to Go From Here
AI adoption is a people problem wearing a technology costume. The organizations that get this right will handle belonging before behavior, culture before tools, and coordination before new instruments.
Start small this month. Name the job-security fear out loud. Ask whether your culture or your tech is in front. Open a channel for complaints. Find your first chair and put their real role in writing.
When you’re ready to put AI to work the way this conversation describes, behind the scenes and in service of the relationship, Virtuous Momentum gives your gift officers AI-powered donor planning, prospecting, and email drafting, so the major-gift treatment can finally reach every donor on your file.
Get a demo of Momentum to see it in action.
Frequently Asked Questions
Why do nonprofit AI rollouts fail?
Most fail at the culture layer, where staff fear about job security makes curiosity impossible. Training and tools can’t fix resistance that’s rooted in an unanswered belonging question.
How do I know if my organization is ready for AI?
Ask whether your tech is in front of your culture or your culture is in front of your tech. If staff already distrust your systems and work around them in spreadsheets, adding AI will widen that gap rather than close it.
What is an AI first chair?
The person on your team who is already experimenting with AI, often informally and without recognition. Identify them, formalize the role in their job description, and meet with them weekly to learn what the rest of the team needs.
Should leaders promise that AI won’t take jobs?
No. Some jobs will change or disappear, and false reassurance erodes trust. The honest path is acknowledging the uncertainty while helping people build skills with the tools now, so they’re positioned to adapt.
How long does this kind of culture change take?
Faster than onboarding new technology. With structured, carefully designed meetings, teams can see meaningful cultural change in under 100 hours of meetings a year.
