AI can write a donor email in seconds. It can research a prospect, analyze giving history, and draft a personalized ask faster than any development officer could type it. None of that is the hard question anymore.
The hard question is whether it should.
That’s the conversation we had with Cherian Koshy on this episode of A Modern Nonprofit Podcast. Cherian is VP at Kindsight and the author of Neurogiving: The Science of Donor Decision-Making, and he’s spent two years running a longitudinal study of more than a thousand donors to find out how people actually feel about AI showing up in the fundraising relationship. The results have held steady across both years of the study, and they point to a simple rule that every nonprofit leader should be repeating to their team.
The line donors have already drawn
Automate the task, not the relationship.
That’s the finding from Cherian and his research partner Nathan Chappell’s donor study, and it’s more useful than most AI advice floating around the nonprofit sector right now. Donors already assume your organization is using AI somewhere. They’re comfortable with it handling back office work: analyzing data, summarizing documents, identifying patterns, drafting a first pass at something a person will edit. What they’re not comfortable with is AI standing in for the human on the other end of the relationship.
That distinction matters because donor trust doesn’t stay contained to one organization. Cherian made a point worth sitting with: when a donor loses trust in one nonprofit’s use of AI, they don’t just stop giving to that organization. They start questioning the sector as a whole. Every nonprofit is carrying some of the reputational weight of every other nonprofit’s decisions about technology.
Where AI actually makes fundraising more human, not less
The best example from the conversation wasn’t about efficiency. It was about a hospital system running thousands of peer to peer fundraising campaigns, the kind where a donor sets up a page in honor of a parent with Alzheimer’s or a family member going through cancer treatment.
Before AI, a staff member manually coded each campaign by cause, and the loop never closed. Nobody ever circled back to say, we saw you did this for your mom. Thank you. With the right use of data and automation, that connection becomes possible at scale. The organization can recognize what a donor did, specifically, and thank them for it, specifically. That’s not more efficient marketing. That’s a more personal relationship, made possible because AI did the unglamorous data work first.
Compare that to the alternative Cherian described from his own inbox: a fundraising email that clearly started as a ChatGPT draft, complete with the tool’s own placeholder language still sitting in the copy. Nobody read it before hitting send. That single mistake tells a donor everything they need to know about how much attention actually went into the ask.
More content is not the same thing as better fundraising
One of the sharper moments in the conversation was about volume. AI makes it easy to produce more: more emails, more appeals, more touches across more channels. The temptation is to treat that as a win.
It isn’t automatically. Cherian and Nathan Chappell describe the risk as a slide toward “spray and pray,” where organizations mistake output for strategy and end up flooding donors with communication that isn’t actually relevant to them. The alternative, what Nathan calls precision philanthropy, uses AI for the part that actually improves quality: cleaning data, segmenting audiences, making sure fewer messages land with more relevance. The goal was never to send more. It was to send the right thing to the right person.
That distinction is worth bringing back to your own systems, not just your fundraising emails. The same principle shows up in financial reporting, in board communication, in grant management. More output isn’t the win. Better targeted output is.
Retire “human in the loop.” Name the human instead.
If your organization’s entire AI policy right now is “we have a human in the loop,” Cherian would tell you that phrase has stopped meaning anything.
It became a comfortable thing to say without ever defining what it actually requires. His recommendation is more specific: a named, accountable person who signs off on AI assisted output, and that person needs the actual subject matter expertise to make the call. Someone with accounting or finance training approves the general ledger coding. Someone with fundraising and donor relations training approves what goes out to a donor. Governance isn’t a vague commitment. It’s a specific person, with the right expertise, taking responsibility for a specific decision.
That’s a useful frame well beyond fundraising. Nonprofit financial management runs into the same trap constantly: a policy that sounds like oversight but was never assigned to an actual person with the training to catch a real problem.
What to do Monday morning
If your team already has an AI subscription and no shared rules around it, Cherian’s advice is to start with a conversation, not a tool selection. Get everyone’s actual feelings about AI on the table. Some people are afraid of it. Some are already moving faster than the organization’s governance can keep up with. Both groups need to be heard before you can build a policy that holds.
From there, decide together what’s appropriate and what isn’t, and look first at the back office work nobody enjoys doing. That’s usually where AI creates the fastest, lowest risk value, long before it should ever touch a donor conversation.
Cherian’s parting thought is worth ending on: most nonprofits are not behind. The organizations chasing every new AI tool out of FOMO are the exception, not the standard you need to hit. Get the culture and the governance right first. The tools will follow.
Want to weigh in on this yourself?
Cherian is currently running the Fundraising Judgment Benchmark, a short study where nonprofit professionals compare AI generated and human generated fundraising recommendations without knowing which is which. It takes about 8 to 10 minutes, and the results will help the sector understand exactly where AI can responsibly assist and where human judgment needs to stay in charge.
Technology changes fast. The fundamentals of running a financially healthy nonprofit haven’t. Clean data, strong internal controls, and people who understand what the numbers are actually telling you still matter more than any single tool. That’s the work The Charity CFO does every day for nonprofits across the country. If your organization has outgrown spreadsheets and everybody figuring it out as they go, learn more about our outsourced accounting and CFO services at thecharitycfo.com.
About Cherian Koshy: Cherian is Vice President at Kindsight and the USA Today bestselling author of Neurogiving: The Science of Donor Decision-Making. Connect with him on LinkedIn or at cheriankoshy.com.
Follow Us Online
🌐 Website: www.thecharitycfo.com
📸 Instagram: @thecharitycfo
📘 Facebook: /thecharitycfo
💼 LinkedIn: / the-charity-cfo
🎵 TikTok: @thecharitycfo
Join our newsletter: https://go.thecharitycfo.com/l/995872/2025-02-24/6ldn1
Download The Charity CFO Financial Blueprint: https://go.thecharitycfo.com/financial-blueprint
Need more than a CFO? Our bookkeeping services offer additional support, so you get full-spectrum financial leadership, all in one place.
