AI Questions Every Nonprofit Board Should Ask
About this resource
Nonprofit boards are being asked to approve AI spending, absorb AI risk, and answer for decisions affecting people the organization exists to serve. This guide gives directors ten questions framed for an organization where the people affected are not customers and cannot take their business elsewhere. It covers mission and program integrity, client data and duty of care, workforce and capability, and stewardship and evidence.
Each question carries what a good answer sounds like and what should concern you. It adds a section on where the freed capacity went, which is where nonprofit measurement differs most from the commercial version, since time saved rarely converts to cost reduction and the board has to decide which kind of return it is asking for. It closes with seven follow-up flags, one question worth not asking in open session, a workable cadence, and the observation that the people carrying the consequence are not in the room.
Boards here carry a duty to people who are not in the room, cannot see the decision, and cannot go elsewhere if it goes badly. Asking who speaks for them is not rhetorical.
How to use it
Ask the same small set consistently and accept an honest "we do not know yet." One addition belongs in a nonprofit context specifically: the people this affects are not in the room, cannot see the decision, and cannot go elsewhere if it goes badly. Asking who speaks for them is the question that most reliably separates a decision the organization can defend from one it will have to explain.
Inside the download
Which part of the mission this serves, and how you would know if it had.
Whether anything sits between a person and the service they came to you for.
Whether funder terms, grant conditions, or client consents already address this.
Which funds are paying and whether the use matches, before an auditor asks.
Common questions
Ten questions across four areas. Mission: which part of the mission this serves, whether anything sits between a person and the service they came for, and what was declined. Duty of care: what client information can and cannot go into these tools, whether funder agreements or client consents address it, and where human review sits. Workforce: what staff actually believe and what would remain if key people left. Stewardship: which funds are paying, and one measurably different outcome.
Three ways. The people affected are not customers and cannot go elsewhere, which raises the duty-of-care weight on any client-facing application. Funder agreements and grant terms may already impose data obligations that predate AI and still bind you. And time saved rarely becomes cost reduction, so the board has to decide whether it is asking for a mission return or a financial one.
It depends entirely on the terms of the restriction, which is why the question belongs in the board conversation before the spending rather than in an audit afterward. A board is entitled to ask which funds are paying and to have management confirm that the use matches the terms.
Free. No email list, no follow-up sequence.
More from the resource library, chosen because they pair well with this one.
Eighteen places to start when you serve people whose privacy you are responsible for, and four places not to.
Ten questions for directors approving something they cannot evaluate, with what a good answer sounds like and what should concern you.
Seven groups, seven different responsibilities, and the assumption that everybody else owns AI.
Tell us a little about yourself and this document is yours. We don't have an email list. We collect this so we know who's finding our resources useful, and that's it.
You're downloading
this document
Your information is kept confidential and never shared with anyone.
