If I had asked you in 1750 whether you used a computer, you would not have pointed at a machine. You would have pointed at a person. “Computer” was a job title. It described someone, usually a woman, who was exceptionally good at mathematics. She sat at a desk with a quill, a ledger, and an extraordinary brain, performing calculations that today take a fraction of a second.
The job existed for over 300 years. Then machines arrived, the role quietly disappeared, and we forgot that a computer was ever a person at all.
I have been thinking about this a lot recently. Because something similar is happening in our sector, only faster.
Your fundraising team is almost certainly using AI already. ChatGPT is drafting thank-you letters. Copilot is summarising meeting notes. Someone in communications is generating campaign images on their lunch break. In most charities, AI has not been introduced. It has simply turned up, like a volunteer nobody quite remembers inviting, and started doing things.
And much of what it is doing is genuinely helpful. Predictive AI is already helping charities identify which supporters to ask, when, and for how much. One UK children’s hospital charity used it to mail fewer donors and raise more income, achieving a 6:1 return on a single appeal. AI is not the enemy. Used well, it is the most exciting thing to happen to fundraising in years.
The tools are in the building. The rules are not.
But here is the uncomfortable part. The research is clear that the vast majority of charities using AI have no formal policy governing how it is used. No red lines. No approval process. No named person responsible.
And that matters, because AI does not understand context. It does not know the difference between a warm relationship and a sensitive one.
Imagine this. A charity uses AI to personalise a lapsed donor reactivation campaign. The AI pulls data from the CRM and generates warm, personal messages at scale. Brilliant, until one lands in the inbox of a recently bereaved supporter, cheerfully referencing their giving “as a couple.”
Nobody intended that. But nobody had told the AI where the lines were. No one had flagged bereavement data as off-limits. No one had built an approval step.
The predictable consequence of tool adoption without governance is not a dramatic scandal. It is a quiet erosion of trust. A donor who stops responding. A complaint that reaches the CEO. A major supporter who pauses their giving and does not say why.
What the research actually shows
McKinsey recently studied nearly 2,000 organisations and found that the ones getting real value from AI are not the ones with the most tools. They are the ones that have worked out where AI acts and where a human reviews. High performers are nearly three times more likely to have redesigned how people and technology work together.
The same research offers reassurance. Most organisations expect their teams to stay the same size or grow. AI is not replacing people. It is freeing them. Just as the human computers gave way to machines, the people moved into roles that needed judgment, creativity, and care. AI takes the repetitive groundwork. Your people do the relational work that actually drives giving.
But none of this works without someone deciding how it should work. And that someone is the board.
Five questions worth asking this quarter
If you are a trustee or a CEO reading this, here are five questions worth putting on the table.
- Is anyone on our team using AI right now, and do we know which tools?
- Do we have a policy covering what data AI can access and who approves its output?
- Have we agreed any red lines, particularly around bereavement, health, and financial hardship data?
- Is there a named person accountable for AI governance?
- If something went wrong tomorrow, could we show our donors and regulators what safeguards were in place?
If the answer to most of those is “no” or “not sure,” you are in good company. But you are also exposed.
The encouraging news is that none of this requires a large budget or a specialist hire. It requires a conversation, a one-page policy, an approval workflow, and a decision about who owns it. It requires leadership.
Three hundred years ago, a computer was a person with a quill. Today, AI can do in seconds what those human computers spent hours calculating. That is remarkable. But the role of humans has not changed. We still set the purpose, the ethics, and the boundaries. AI does not reduce that responsibility. It raises the stakes.
Your team is already using AI. The only question is whether anyone has agreed the rules.
Darren Richards is the founder of Charity AI Partners and advises charity CEOs, directors and trustees on AI-ready fundraising strategy. He has over twenty years in international fundraising, is the author of The Plane Parable, and holds a certification in AI governance for trustees and directors. Writing in a personal capacity. @CharityAIGuy