By Sterrin Bird, CFRE  |  Wingspan Philanthropy

The consensus arrived fast and it arrived confident. Artificial intelligence is coming for knowledge work. The analysts go first, then the writers, then anyone whose job can be described as reading things and producing other things. Fundraising shows up on most of these lists. Prospect research, proposal drafting, donor communications, portfolio management. All of it looks automatable, because most of it is.

I think philanthropy is about to run the other way.

And I think the first casualty will not be the fundraiser. It will be personalization itself.

Not because our sector is special. Because of what happens when you subtract everything a machine can do and look honestly at what remains.

Start by conceding the ninety percent

Let me give away more than most consultants will.

AI will do prospect research better than any analyst on your team. It will read the 990s, cross-reference the real estate records, surface the board affiliations, and flag the capacity you missed. It will draft the proposal. It will prep the meeting. It will write the follow-up note in your voice, correctly, in four seconds, and it will remember every detail of the last conversation better than you did.

None of that is a threat to make peace with. It is a gift. Most gift officers spend the majority of their week on work that does not require a human being, and they spend it badly, because they were hired for something else entirely.

So hand it over. All of it.

Now look at what is left.

What is left is a narrower job, and a harder one

Strip out the research, the drafting, the preparation, the administrative sediment, and the role reduces to two things. Time in the presence of another person. And judgment about what to do with that time.

That is a smaller job than the one most gift officers hold today. It is also considerably more difficult.

Here is the uncomfortable part. A great many people currently succeeding in this profession are succeeding at the tasks that are about to disappear. They are organized. They are thorough. Their files are immaculate and their reports are on time. Those were real virtues in a world where the surrounding work was expensive. In a world where it costs nothing, they are table stakes.

So this is not the rise of the gift officer. It is the rise of some gift officers. The sorting will be fast and it will not be gentle.

The danger is not replacement

The actual risk is subtler than the one everyone is arguing about, and it is already underway.

AI will make mediocre gift officers look productive.

Touches climb. Portfolio coverage improves. Response rates hold steady. The dashboard has never looked better, and every metric leadership tracks will confirm that the program is healthy. Three years will pass before anyone notices that nothing deepened. That no relationship moved from cordial to committed. That the pipeline is wide and utterly shallow.

I have spent a career objecting to expensive activity that never moved a donor. The eighty thousand dollar feasibility study. The audit that produced a binder and no gifts. This is the same disease with a faster engine. Volume was always the easiest thing to measure and the least worth measuring, and we are about to make volume nearly free.

Now turn it around and look at the donor

Half the sector has not thought this through.

Your donors have the same tools you do.

Consider what that does to personalization, which we have spent fifteen years treating as the summit of donor communication. The perfectly tailored letter. The message that references her daughter’s graduation and his interest in watershed restoration and the trip they took in 2019. We built entire technology stacks to make that possible at scale.

That signal is about to go to zero. Not because the letters get worse. Because they get uniformly excellent, and a sophisticated donor will correctly assume a machine wrote it, since usually one did. Effort that can be manufactured infinitely stops communicating effort. It communicates nothing at all.

What survives is what remains expensive to fake.

Showing up in person when it was inconvenient. Remembering the thing that was never written down anywhere, because you were actually listening when she said it. Sitting with a donor through a hard year and asking for nothing. These are costly signals, and costly signals are the only ones that carry information.

Where data ends, devotion begins

There is one more thing a machine cannot do, and it is the one I would build a career on.

An AI optimizing for donor engagement will never tell a donor that her idea is wrong for the organization.

It cannot. It is built to please, trained on approval, tuned toward the response that produces satisfaction. Ask it to counsel a donor away from the restricted gift that would distort your programs, and it will find the version of that message with the least friction in it. It will smooth the objection until the objection is gone.

It cannot risk the relationship, because it does not understand that the willingness to risk the relationship is what makes the relationship real.

That is the whole profession, in one sentence. The gift officers who matter are the ones who will say the true and unwelcome thing to a person who could walk away. Donors know the difference. They have always known the difference. They have simply had very few opportunities to encounter it.

The machines take the flattery.

Humans keep the friction.

It turns out the friction was the gift all along.