Practical AI

Use AI Without Losing the Human Judgment Behind Donor Trust

By Mission ArcOctober 5, 20262 min read

---

AI can help a fundraising team get from a blank page to a first draft. It can also produce confident language that is inaccurate, generic, or unsuitable for the relationship involved.

A useful workflow makes those limitations visible. Start with a specific task, define the information that may be used, and assign responsibility for reviewing the result.

Choose a bounded task.

“Help us raise more money” is too broad to guide a dependable process. “Create three outline options for an impact update using these approved program facts” is more specific.

Other starting points might include organizing a nonconfidential planning agenda or adapting approved public copy for a different format. Choose work where the output can be checked against a clear source.

Decide what information belongs in the tool.

Before using a tool, follow your organization’s policies and review its approved uses. Do not assume that a tool is appropriate for identifiable donor information simply because it is convenient.

For a first workflow, use public or otherwise approved material. Remove unnecessary personal details. If you cannot determine whether information is suitable to share, resolve that question before uploading it.

Give the tool boundaries.

Tell it which facts to use, who the audience is, what format you need, and what it must not invent. A useful instruction is: “Use only the supplied facts. Flag missing information instead of filling it in.”

That instruction does not make the output automatically reliable. It establishes a standard the human reviewer can check.

Make review part of the task.

Assign a person to compare the draft with its sources. Check names, dates, amounts, descriptions of impact, and any claim that could influence a donor’s understanding.

Then review the tone. Does the message respect the people whose stories are included? Does it sound like your organization? Does it imply a promise you are not able to make?

Keep the relationship human.

AI should not decide what a donor values, invent a personal history, or send a sensitive message without a person’s approval. The team remains responsible for the communication and its consequences.

Document the approved workflow, including what goes into the tool and who checks what comes out. If the task changes, revisit those boundaries.

Begin with one useful experiment.

Select one low-sensitivity task, use approved source material, review the result carefully, and decide whether the process helps. Keep what is useful and revise what is not.

The goal is practical support for the work, with accountability still in human hands.