.png)
AI Can Help Us Ask Better. It Still Can't Tell Us What Matters.
Reflections on intentionality, curiosity, and human judgment after AHP International
"Validate!"
By the end of my session at the AHP International Conference, that single word had become something of a refrain, with nearly the entire room calling it out.
And that was exactly the point. During "Dear AI: Tell Me What I Meant to Say — A Fundraiser's Guide to Better Asking," I encouraged participants to think beyond what AI could produce and consider what they needed to question, verify, and evaluate before using it. One attendee even referenced the moment in a LinkedIn comment afterward, remembering that she had called out "Validate!" during the session.
What was especially encouraging was hearing from attendees who had already begun applying what they'd learned as early as Monday, when they returned to work. That feedback meant a great deal to me. My goal wasn't simply to introduce another way to write prompts. It was to help fundraisers approach AI with greater confidence, structure, and intentionality. I had developed CROST-V™ as a framework to support that thinking, and AHP gave me an opportunity to share it with fundraising professionals.
But what stayed with me went beyond the presentation itself. Some fundraisers were already comfortable experimenting with AI. Others were still exploring where to begin. And for some, there was understandable hesitation about how to use it responsibly, particularly when it came to sensitive information, accuracy, and donor privacy. The questions weren't necessarily surprising. Many reflected concerns we've been hearing as AI becomes more accessible.
What I took away from those conversations was a reminder that learning to use AI is only part of the challenge. As we become more comfortable using these tools, are we giving equal attention to the judgment, intentionality, and responsibility that come with them? It's a question I've been thinking about for some time, and the conversations at AHP reinforced why it's worth exploring.
The Better Prompt Isn't the Point
I've spent considerable time exploring how the way we communicate with AI influences what it produces. A vague request can generate a generic response. Providing clearer context, purpose, and expectations can make the result substantially more useful. That's part of what makes thoughtful prompting valuable. But better prompting is only the beginning.
Consider how we might use AI to develop a donor communication. We can ask it to refine the tone, suggest alternative language, organize information, or tailor a message for a particular audience. In moments, we may have something polished, persuasive, and seemingly ready to send. But should we send it?
Does it accurately represent the organization and its mission? Does it reflect what we actually know about the donor? Is it appropriate for the relationship? Have we verified the information? And does the message say what we genuinely intend to communicate? Those questions require more than a well-constructed prompt. They require professional judgment.
One of the exercises I shared at AHP explored this distinction. We looked at different versions of a fundraising appeal developed from the same source material and considered which would be most appropriate for its intended audience. The point wasn't necessarily to identify the version that sounded the most compelling. It was to recognize that the best-written response isn't always the best response for the purpose or the audience. That distinction matters whether we're preparing a donor appeal, analyzing information, or developing a strategy.
Validation, then, is about more than checking whether AI invented a statistic or misrepresented a source. It's also about evaluating whether what we've produced is accurate, appropriate, and genuinely useful. A better prompt can help us produce a better answer. But we remain responsible for deciding whether it's the right one.
Are We Becoming More Effective, or Just More Efficient?
One of AI's most compelling promises is the ability to give us back time. And in fundraising, time is a valuable resource. We're balancing donor relationships, proposals, stewardship, prospect development, communications, reporting, and the many operational demands that support our work.
If AI can help us organize information, prepare for meetings, or develop a first draft more efficiently, that's a meaningful opportunity. But I wonder whether we're asking enough questions about what happens after we save that time.
The 2026 Nonprofit AI Adoption Report, presented by Virtuous and Fundraising.AI, offers an interesting perspective. In a December 2025 survey of 346 nonprofit organizations, 92% reported using AI in some capacity, while only 7% reported major improvements in their organization's ability to achieve its mission. These are self-reported findings from participating organizations, not a representative measure of the entire nonprofit sector. They also don't establish whether any particular use of AI improves fundraising outcomes.
Still, the distinction between adoption and impact is worth considering. We can become faster at writing emails without necessarily becoming better communicators. We can produce more donor research without understanding donors more deeply. And we can complete more activities without necessarily improving the quality of our relationships.
That doesn't diminish the value of efficiency. It challenges us to think more intentionally about how we use it. If AI helps us spend less time preparing a document, could we spend more time having meaningful conversations? If it helps us organize information more effectively, could we use that information to ask better questions? And if it gives us additional capacity, how might we reinvest that capacity in the people and relationships at the center of our work?
Perhaps the opportunity isn't simply to do more. It's to create more space to do what matters well.
Curiosity: An Opportunity We Shouldn't Overlook
There's another dimension of AI that I think deserves more attention in fundraising.
We often discuss what AI can produce. But what about what it can help us explore?
Think about the conversations we have with donors. Someone shares an experience, an interest, a book they've been reading, a project they've worked on, or something personally meaningful that we may know very little about. Sometimes those subjects are directly connected to the mission. Other times, they're simply part of who that person is. And those moments can open the door to something much more meaningful than another fundraising conversation.
AI can help us explore an unfamiliar topic, better understand a concept, consider different perspectives, or prepare thoughtful questions that might not otherwise occur to us. Used appropriately, it can help us arrive at a conversation better informed and more prepared to engage. But here's the distinction I think matters.
AI can help us ask more thoughtful questions. Genuine curiosity is what makes us interested in the answers.
There's a difference between asking questions because we've identified a potential giving opportunity and asking because we're genuinely interested in understanding someone. Experienced fundraisers know that relationships rarely develop according to a perfectly structured cultivation plan. Sometimes the most meaningful conversations happen when we move beyond the agenda we brought into the room. We learn something unexpected. We discover a shared interest. We understand an experience that has shaped someone's perspective.
And sometimes, what we learn tells us that the next step isn't a solicitation or another formal engagement opportunity. It's simply continuing the conversation. AI can support our preparation for those moments, but it cannot guarantee them. It can't make our interest genuine. And it shouldn't become a substitute for listening.
This also makes me think about how we document relationships. Our constituent relationship management systems (CRMs) serve an essential purpose. They help us record interactions, preserve institutional knowledge, coordinate stewardship, and support continuity when multiple colleagues are involved.
But a meeting recorded in the CRM doesn't necessarily tell us whether a donor felt heard. An email exchange doesn't automatically reveal whether trust developed. And a contact report, however thoughtful, may capture only part of what made a conversation meaningful. That doesn't diminish the importance of documenting activity. It reminds us that recording what happened isn't always the same as understanding what developed.
And perhaps that's another opportunity AI creates for us—not simply to increase the number of interactions we have, but to help us bring greater curiosity and preparation to the interactions that matter.
The Judgment Behind the Recommendation
In fundraising, we're accustomed to working with information that helps guide decisions. Giving histories, engagement patterns, prospect research, and capacity assessments all provide valuable insight. AI can help us organize and analyze that information, sometimes revealing patterns or possibilities worth exploring. But identifying a potential opportunity isn't the same as understanding a donor's readiness.
I've encountered situations where a donor's capacity suggested the possibility of a larger gift, but what we understood about their interests, priorities, or readiness suggested a different approach. I've also seen strategies that made sense on paper but needed additional context before we could act on them appropriately. Those situations aren't unique to AI. They're part of the professional judgment fundraisers have always exercised.
What's changing is how quickly AI can help us generate recommendations, summaries, and potential strategies. A recommendation may be reasonable based on the information available. But what might be missing? And what do we still need to understand directly from the donor before deciding what happens next? Those questions become especially important when AI-generated outputs appear confident and complete.
The Association of Fundraising Professionals' October 2026 guidance on ethical AI use reinforces the importance of human review, responsible handling of donor information, and professional accountability. That means verifying the information AI produces, understanding the limitations of the tools we're using, and protecting the data entrusted to our organizations. It also means recognizing that donor information isn't ours to use indiscriminately simply because it might help generate a more personalized communication or recommendation.
For me, the opportunity is to use AI to inform our thinking without allowing the speed or apparent confidence of its output to substitute for the judgment a donor relationship deserves.
What Does the Donor Experience?
There's another perspective worth considering: how donors feel about the increasing use of AI in philanthropy. In their 2026 Donor Perceptions of AI study, Nathan Chappell and Cherian Koshy found that 77% of the 1,020 U.S. donors surveyed said transparency about charitable organizations' use of AI was very important. The study also found that greater familiarity with AI did not necessarily translate into greater comfort with its use by charities.
These findings reflect the donors surveyed, not necessarily the views of every donor. But they raise an interesting question. As we become more comfortable using AI, are we giving equal consideration to how donors experience its use? And I don't think every application of AI should be treated the same way.
Using an approved tool to brainstorm questions before a meeting is different from allowing an automated system to communicate with a donor as though it were a person. Likewise, using AI to help prepare a thoughtful message is different from sharing sensitive personal information in ways the donor wouldn't reasonably expect. Those distinctions matter because trust isn't simply about whether we use AI. It's also about how we use it, what information we share, and whether our practices respect the people behind that information.
The opportunity isn't to make every communication appear more personal. It's to use technology in ways that support more thoughtful, authentic engagement while remaining mindful of the trust donors place in our organizations.
A Reminder of Why People Give
Perhaps that's why Miguel Cervantes' keynote at AHP International stayed with me. His deeply personal story about his daughter, his family's experience with loss, and their commitment to supporting CURE Epilepsy was inspiring. It wasn't a presentation about AI. But after so many conversations about technology, strategy, growth, and how we can improve our fundraising practices, his story was a powerful reminder of why people become involved in philanthropy in the first place.
Sometimes it's gratitude. Sometimes it's hope. Sometimes it's a personal experience, a loss, or a commitment to making something better for someone else. And sometimes the motivations behind a philanthropic relationship are far more personal and complex than we could ever understand from a giving history or an engagement score.
That's what makes curiosity so important. We can't assume we know what motivates someone simply because we've analyzed their behavior. We have to be willing to listen, to ask questions, and to allow people to share what matters to them. And we need to respect that they may choose not to share everything.
For me, Miguel's keynote reinforced something that can get lost when we're focused on tools, processes, and performance measures. Philanthropy is deeply human.
Technology can help us prepare for conversations, understand unfamiliar subjects, and recognize opportunities to connect. But the meaning behind someone's involvement in a cause emerges through their experiences, values, and relationships. Our responsibility is to approach those relationships with care.
Beyond the Better Ask
When I think back on AHP, I'm encouraged by the conversations, the willingness of fundraisers to explore new approaches, and the stories I've heard from those already putting what they learned into practice. But I also keep coming back to what we're ultimately trying to accomplish. We have opportunities to use AI to become better prepared, explore unfamiliar ideas, ask more thoughtful questions, and perhaps create more space for the relationships that are central to our work. And that's exciting.
But using AI well isn't simply about getting a better response. It's about knowing what to do with that response—and recognizing when we need to ask another question, seek more context, or listen more carefully. For me, that's where intentionality becomes especially important. Because the same questions extend beyond AI.
We have increasingly sophisticated tools for understanding donor behavior, identifying opportunities, and managing relationships. Yet recognizing potential is not the same as knowing what a relationship needs next. And perhaps that's what I've been thinking about most since AHP. Not just how we become better at asking AI for what we need, but how we become more thoughtful about the decisions and relationships those answers are meant to support.
As our tools become more capable of helping us identify opportunities, are we being equally intentional about what happens next?
References & Suggested Readings
Virtuous & Fundraising.AI. (2026). The 2026 Nonprofit AI Adoption Report. An industry-sponsored benchmark survey of 346 nonprofit organizations examining AI adoption, reported impact, organizational readiness, and the distinction between productivity improvements and strategic value. Explore the report
Alotto, J., & Ehrlich, S. (2026). AI for Good: Building an Ethical Approach to Artificial Intelligence. Association of Fundraising Professionals. October 15, 2026. Professional guidance on responsible AI adoption, donor privacy, validation, transparency, and human accountability in fundraising. Read the AFP article
Chappell, N., & Koshy, C. (2026). Donor Perceptions of AI 2026: Three Years of Listening to Donors. Fundraising.AI. Independent survey research exploring donor familiarity with AI, expectations about transparency, concerns about AI-generated engagement, and charitable giving intentions. Read the study
National Institute of Standards and Technology. (2024). Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile (NIST AI 600-1). A technical resource examining generative AI risks and approaches to identifying, assessing, and managing them. Relevant to responsible use and validation, although not specifically focused on fundraising. Read the NIST publication
Disclaimer
The views and opinions expressed in this article are solely those of the author and do not reflect the official policy, position, or endorsement of any organization with which the author may be affiliated. This article is intended for educational and informational purposes only. Professional experiences and observations reflect the author's perspective and are not presented as research findings. Readers are encouraged to consider their own organizational context, policies, and professional judgment when applying these ideas.
Comments