It usually happens somewhere between the financial report and new business. A board member, maybe the one who works in tech, maybe the one who just read a worrying headline, leans forward and asks: "Are we behind on AI?" And because you run communications, or development, or the whole organization, everyone in the room waits for you to answer.
If that question hasn't landed yet, it's probably headed for one of your fall meetings. Board members hear about AI constantly, and they're not wrong to bring it up: supporters now ask ChatGPT and Perplexity which organizations to give to, volunteer with, and trust, a shift we've written about in detail in how AI search is changing the way people find nonprofits.
Before you answer, it helps to notice what's actually being asked. "Are we behind on AI?" sounds like a question about tools, but underneath it's a question about relevance and risk, which are the two things boards are responsible for. Answer it at that level and the conversation usually goes well. Answer it with a list of software you may or may not adopt and it usually goes badly.
So this post is a practical guide to how to answer board questions about AI at your nonprofit: the translation, the five-minute prep, sample language for either outcome, and a one-page briefing template you can copy into your next board packet.
Translate the question before you answer it
"Are we behind?" is hard to answer as asked, partly because it never specifies behind whom. It becomes workable once you break it into three questions you can actually answer:
1. Can supporters find us through AI assistants?
When someone asks an AI assistant for "organizations doing food security work in our county," do you appear? This is a visibility question, and it has a technical answer. AI assistants read the raw structure of your website, and the practice of making your content readable and citable by them has a name: generative engine optimization, or GEO. Think of it as search engine optimization for a world where the "search engine" writes sentences instead of listing links.
2. Is what AI says about us accurate?
Assistants describe your organization whether or not you participate. They summarize your programs, your history, sometimes your finances. Ask one about your organization tonight and read the answer carefully, because that's what a prospective donor sees.
3. Are we using AI responsibly internally?
This one isn't about visibility at all. It's about staff use, disclosure, and data handling, which makes it a policy question and means it deserves a policy answer. If your organization doesn't have one yet, our guide to AI content governance for nonprofits walks through what a right-sized policy covers.
The useful thing about this breakdown is that the first two questions can be diagnosed this week, and the third is a document your board can help you shape. Framed that way, "are we behind" stops being a vague anxiety and becomes an ordinary agenda item.
The five-minute prep: bring a score, not a feeling
The weakest possible answer to a board is "I think we're okay." It may well be true, but a feeling can't be tracked, budgeted against, or revisited next quarter, so before the meeting, run a graded check instead. Our AI Visibility Scorecard reads your website the way AI assistants do and returns a grade in about two minutes. It's free, and the grade is designed to be shown to non-technical people, which makes it a natural fit for a board packet.
Be precise with the board about what the score covers. The Scorecard answers question one (can AI find and read us) and partially answers question two, because a site AI can't read is a site AI will describe from other, older sources. It does not answer question three; no scan can tell you whether your staff's use of AI matches your values. That remains the policy conversation, and saying so out loud is part of what makes the briefing credible.
If the grade surfaces problems you want examined properly, a deeper AI visibility audit is the logical next step. For the board meeting itself, though, the grade is enough.
Sample language for every outcome
Boards reward a diagnosis with a plan attached. They are suspicious of pure reassurance, and they dislike surprise. Whatever your grade, the structure that works is the same: here's what we measured, here's what it means, here's what we're doing.
If the grade is strong:
"We ran a graded check of how AI assistants see us. We scored well — our site is readable to them and the basics are in place. Two things we're watching: keeping our content fresh enough that assistants cite current information, and putting a simple internal AI-use policy in front of you next quarter."
If the grade is weak:
"We ran a graded check of how AI assistants see us, and we found a specific gap: our site is hard for them to read, which means when supporters ask an assistant about our work, the answer comes from old sources or not at all. Here's the three-step plan to close it, what each step costs, and when we'll re-run the check."
Notice what the weak-grade version doesn't do: it doesn't apologize, it doesn't catastrophize, and it doesn't ask for a large budget on the spot. It names the gap and attaches a plan, which is what competence looks like from a board seat. In our experience, the executive who found the problem and brought a plan leaves the meeting with more of the board's confidence than the one who never looked.
The point that lands hardest
If you make one substantive point beyond the grade, make this one.
AI assistants don't work from your website alone. They learn from archives and from old crawls (a crawl being the automated pass a piece of software makes through your site to read and store its pages), and that stored copy might be years old. This is why assistants confidently repeat stale facts about organizations: program names retired two rebrands ago, events that no longer run, pages that now return errors.
Here's the governance framing, in governance language: there is a public record about our organization that we did not write, cannot edit directly, and that donors are already consulting. You can't correct that record with a phone call or a press release. The only correction mechanism is publishing fresh, well-structured, machine-readable content and keeping it that way.
Boards understand records risk. Put the situation in those terms and the room stops treating AI as a marketing curiosity and starts treating it as something the organization is accountable for, which it is.
The one-page briefing template
Here's the whole briefing, ready to copy. It fits on one page on purpose — a diagnosis and a plan that can't fit on a page usually isn't finished yet.
AI readiness briefing: [Organization name], [Date]
Our score: [Grade] on the AI Visibility Scorecard, run [date].
What it means in plain words: [e.g., "AI assistants can find and read our site, but our program pages lack the structure assistants cite from" (one or two sentences, no jargon).]
What AI currently says about us: [One sentence: accurate / partly stale / wrong in these specifics.]
Three next steps:
- [e.g., Fix the readability gaps the scorecard flagged (staff time, est. X hours)]
- [e.g., Refresh the three pages assistants cite most often (staff time)]
- [e.g., Draft an internal AI-use policy for board review next quarter (staff time)]
What this costs: Mostly staff time. [Name any outside cost honestly, or state there is none yet.]
How we'll report back: Re-run the graded check in [month]; grade and progress on the three steps in the next board packet.
Adapt it freely, because the shape is what matters: a score, a plain-words meaning, numbered steps, honest costs, and a date you'll report back. That last line quietly turns a one-time scare into a managed, recurring metric, which is exactly where a board wants this topic to live.
Run the check before the agenda goes out
Very little of this requires a big project up front. The difference between a tense AI conversation and a good one is about two minutes of preparation, so run the AI Visibility Scorecard before your next board meeting. It's free, it takes about two minutes, and it returns a grade you can drop straight into the briefing above. Then, if the question comes up somewhere between the financial report and new business, you'll have something concrete to say.
