Fundraising.AI Global Summit 2026
People Shaping AI in Nonprofit Fundraising in 2026
Our living notebook for the Fundraising.AI Global Summit 2026: ten people whose ideas we follow, and what those ideas mean for grant work.
Mira Pape & Tumelo Peane · October 14, 2026 · updated October 15, 2026 · 28 min read
We’ve been following the Fundraising.AI Global Summit since 2024, and it has become one of the events we look forward to most.
We first found the Summit through Mallory Erickson, whose work we already loved for its focus on behavior, confidence and the human side of fundraising. That perspective felt refreshingly different from a lot of the fundraising education we were seeing in Germany at the time. So when she helped launch the first Fundraising.AI Summit, we were immediately interested. Three years later, it has become one of our favorite places to learn about AI in the nonprofit sector.
For the last couple of years, we kept saying that we should properly document our notes and thoughts from the sessions. And then the Summit happened, our normal work happened, and somehow those notes never turned into anything useful.
This year, we finally blocked out the time.
We had our notebooks ready, the Summit schedule open, and a list of people whose work we wanted to follow particularly closely. We published this article before the Summit and are updating it as we go.
Our own work is mainly in European institutional funding, and there is unfortunately no dedicated grant-writing session this year. So alongside the Summit takeaways, we’ll keep asking:
What does this mean for grant work?
We’ll translate the ideas into funding research, project development, proposal writing, review and knowledge management, and add a European perspective where it matters.
Think of this as our living Summit notebook, just a little more organized.
Live Summit notes
Day 1: What’s our four-minute mile?
For a long time, running a mile in under four minutes was seen as almost impossible. Then, in 1954, Roger Bannister did it. Just 46 days later, John Landy broke his record.
People didn’t suddenly become better runners overnight. But someone had shown that the barrier could be broken.
Nathan Chappell and Mallory Erickson brought up the four-minute mile in their opening keynote, The State of Fundraising AI, and with a background in sports and sports nonprofits, Tumelo was hooked straight away.
So much of the AI conversation is about efficiency: writing proposals faster, automating admin, saving a few hours here and there. Useful, yes. But what if we’re aiming too low?
- What if a two-person nonprofit had the reach of a team of twenty?
- What if every great project found the funding it needed?
- What if nonprofits could spend nearly all their time on their mission, not on administration?
- What if our biggest challenge was figuring out what came next, because we’d actually achieved our mission?
AI isn’t going to magically solve the world’s problems. But with so much changing around us, it’s worth questioning some of the limitations we’ve come to accept. One line from the keynote really stayed with us:
“We already have this level of fear. Why can’t we have this level of creativity?”
So we’re curious: what’s your four-minute mile? What would you love to see become possible in your organization, or in the nonprofit sector, even if it sounds completely unrealistic today?
More from the opening keynote: Mallory Erickson on moving beyond the easy use cases and Nathan Chappell on spending time without AI.
Steal Their Best Ideas
10 practical experiments inspired by the people shaping AI in nonprofit fundraising.
Why this idea matters
Mallory Erickson’s work increasingly focuses on fundraiser readiness rather than simply teaching more tools or tactics. That is one reason we keep coming back to her perspective on AI.
At the 2025 Fundraising.AI Summit, one of her strongest points was that it is easy to become reactive around AI. There is always another tool, another promise, another thing that suddenly feels urgent. Her response was simple: slow down, identify the biggest challenges first, and only then decide what kind of solution actually makes sense.
At the end of the opening session, she turned that into a practical exercise. Before participants even looked at the Summit agenda, she asked them to name the three biggest pain points in their organization or leadership. Only then should they decide what was worth paying attention to.
What it means for your grant work
For grant professionals, that is a very useful shift.
Instead of starting with “Which AI tool should we use?”, start with “Where are we actually losing time, quality or consistency?”
Maybe the biggest issue is not writing at all. Maybe it is identifying strong-fit funders, collecting information internally, reusing organizational knowledge, or adapting the same project logic to different programs.
Once the real friction is clear, you can decide whether AI is part of the solution and what kind of support would actually help.
What we’re listening for at the Summit
This year, we’ll be listening for how Mallory develops that idea of agency and proactive decision-making now that AI tools are becoming more powerful and more embedded in everyday nonprofit work.
What we heard at the Summit
Mallory and Nathan opened the Summit with a bang. Again.
Mallory talked about the hesitation many of us still feel around AI. Maybe our friends and family don’t like it. Maybe we worry that by using AI, we’re somehow “participating in the downfall of humanity.”
These are real concerns, and we should be asking questions about the consequences of using AI. But one of our biggest takeaways from her session was this:
As nonprofits, it is part of our responsibility to move beyond the easy AI use cases.
Our responsibility doesn’t end with using AI responsibly. It’s also about being curious enough to explore what else might be possible.
Of course we need to think about data protection, ethics and which tools we can safely use. But we sometimes spend so much time discussing which tools we should or shouldn’t use that we barely get around to asking what we could actually achieve with them.
For grant work, that means looking past “write the proposal faster” and asking bigger questions, like what it would take for every great project to find the funding it needs.
Profile
Kaz McGrath
Main idea
Stop trying to write the perfect prompt. Explain the task like you would to a colleague.
Who she is
Founder & CEO, PLAI – Purpose Led AI
What she’s known for
Mission-led AI adoption, AI fluency, experiential learning, and making AI feel practical and approachable for nonprofit teams.
Where to find her
Fundraising.AI Summit 2026
Put the Tool Down! Human Skills that Ensure Impact from AI
with Stephen Silveira
Kaz is one of this year’s Summit hosts, together with Mallory Erickson and Tara Quarrie, and appears across several AI Sparks sessions. The session above is her main Day 2 session.
Why this idea matters
One of our favorite sessions from last year was Kaz’s short session on prompting.
Not because she introduced a complicated new framework. Almost the opposite.
Her point was that people often get stuck trying to type the perfect prompt when they already know how to explain work to another human. Instead, she asked people to imagine giving the task to a colleague.
- What are you trying to achieve?
- What information and resources would you give them?
- What does a great result look like?
- What are they likely to get wrong?
- What examples would help them understand?
That is essentially the prompt.
She also encouraged people to use voice, rant, think out loud and have a real back-and-forth conversation with AI rather than treating every interaction like a one-shot command.
We love this because it makes working with AI much less technical. The skill is not finding the perfect words. It is learning to brief, delegate, explain and give feedback well.
What it means for your grant work
Grant professionals already know what a strong application, research brief or project concept should look like. The challenge is transferring that knowledge into the AI interaction.
Kaz actually demonstrated this with a grant-research example: instead of asking AI to write the brief, she asked it to guide her through the strategic thinking and questions needed to build one, then used feedback to push the result deeper.
For us, that is the useful shift:
Do not ask AI to do the work before you have explained how good work gets done.
What we’re listening for at the Summit
Kaz’s session this year is Put the Tool Down! Human Skills that Ensure Impact from AI.
We’ll be listening for which human skills she now sees as most important, and whether the same challenge she highlighted last year still shows up: people knowing what good work looks like, but struggling to translate that knowledge into a useful conversation with AI.
Steal Their Best Ideas10 practical experiments inspired by the people shaping AI in nonprofit fundraising.
Get the workbook →Why this idea matters
What we like about Tara’s work is that she makes “responsible AI” much more practical.
It is easy to say that an organization should use AI ethically. It is much harder to decide what that actually means when someone wants to use AI for a specific task next Tuesday.
Her TRUST™ Framework helps move that conversation from general principles to concrete questions about transparency, consent, unintended consequences, staff and stakeholder voice, and human oversight.
The broader idea we take from her work is this:
AI readiness is not one big yes-or-no question. It depends on the use case.
An organization may be perfectly comfortable using AI for one task and not ready at all for another.
What it means for your grant work
That distinction matters a lot in grant work. Using AI to summarize a publicly available funding guideline is very different from asking it to work with an unpublished project concept, a detailed budget, partner information or internal evaluation data.
So before introducing AI into a grant workflow, ask:
- What information does this task require?
- Where does that information come from?
- Can we put it into the tool we are using?
- Who should be able to access the result?
- And who is responsible for checking it?
The useful question is therefore not simply “Are we allowed to use AI for grant writing?” It is:
“Are we ready to use AI for this particular part of the grant process?”
What we’re listening for at the Summit
This year, Tara is presenting The TRUST™ Framework.
We’ll be listening especially for how she moves from ethical principles to real implementation decisions, and which risks she now sees development teams struggling with most as AI becomes more deeply embedded in everyday fundraising work.
Profile
Scott Rosenkrans
Main idea
Predictive, generative and agentic AI are three different jobs.
Who he is
Head of AI Growth, EverTrue
Co-Author, Nonprofit AI
What he’s known for
Practical AI adoption in fundraising, translating different forms of AI into concrete use cases, and helping teams move from curiosity to operational use. His recent work spans predictive AI, generative AI, agents, automation and adoption.
Where to find him
Fundraising.AI Summit 2026
Overwhelmed by AI? Slow Down and Lead Well
with Mallory Erickson
Why this idea matters
One thing we find useful in Scott’s work is how clearly he separates different types of AI instead of treating “AI” as one thing.
- Predictive AI helps you decide where to focus.
- Generative AI helps you create or shape content.
- Agentic AI can take actions based on information, rules or goals.
That distinction matters because a lot of organizations still use the word AI when they really mean one very specific thing: generating text.
Scott’s broader work is useful precisely because it moves the conversation away from “What can ChatGPT write?” and toward which kind of AI is actually suited to which kind of task.
What it means for your grant work
This distinction maps surprisingly well onto grant work.
- Predictive AI could help with funder fit, prioritization or deciding which opportunities deserve attention first.
- Generative AI can help structure information, draft text, adapt content and support review.
- Agentic AI could monitor funding opportunities, track changes, flag deadlines or move information through a recurring workflow.
Most grant professionals currently experience AI mainly through the generative layer. That is useful, but it is only one part of what AI can do.
Thinking in these three categories makes it much easier to ask:
What kind of job are we actually trying to get AI to do here?
What we’re listening for at the Summit
Scott is speaking with Mallory in Overwhelmed by AI? Slow Down and Lead Well.
We’ll be listening for whether his focus this year is still mainly on helping teams distinguish between different AI capabilities, or whether the harder question has now become how organizations decide which capabilities they are actually ready to use.
Why this idea matters
Nathan has been talking about AI in philanthropy for years, long before generative AI became part of everyday nonprofit work.
One of the ideas we keep coming back to in his work is human first, AI forward.
It is a useful counterpoint to the more common “AI first” mindset. The idea is not to reject the technology, but to stay clear about what the technology is there to support. That matters especially in a sector built on trust.
Nathan’s broader work repeatedly makes the point that better AI adoption is not just about giving more people access to tools. It also requires organizations to rethink workflows, guardrails and the way people actually work together.
What it means for your grant work
A grant application is not just a piece of writing.
It has to convince a funder that your organization understands the problem, knows the people or communities it wants to work with, has realistic plans, and can actually deliver what it is proposing.
That is why we think human first, AI forward is such a useful principle for grant work. AI can help with research, structure, drafting, review and feedback. But the final application still needs to be grounded in your organization’s actual experience, decisions, relationships and ability to deliver.
If AI helps you express that more clearly, great. If it starts replacing the thinking behind it, that is where the quality of the application can quickly drop.
What we’re listening for at the Summit
Nathan is opening the Summit again this year with Mallory and is also opening Day 2 with Gabe Cooper on N1 Philanthropy.
We’ll be listening for how his “human first, AI forward” idea develops as AI becomes more autonomous, and where he now draws the line between work AI should take over and work that should remain fundamentally human.
What we heard at the Summit
For every hour you spend with AI, spend one without it.
Nathan basically reminded us to go touch grass. Go outside. See some art. Make some art. Spend time doing things that have absolutely nothing to do with AI.
It got Mira thinking about how AI has actually made her more interested in doing things offline. She loves redesigning spaces and starting creative projects, but has always been prone to analysis paralysis: research, rethink, look for more inspiration, and sometimes never start. Now she can describe what she has in mind, see what a room could look like or work out which materials an art project needs. It doesn’t make everything perfect, but it gives her enough confidence to get started.
There’s another side to it, though. AI can be ridiculously addictive when you enjoy creating things. There’s always another idea to test. Having a family forces her to close the laptop, go outside and make things with her hands, in a world where nobody cares what the latest AI model can do.
We don’t want AI to replace the things we love doing away from the computer. We want it to help us do more of them.
The same goes for grant work: if AI takes over some of the desk work, the time it frees up is best spent on what no tool can do for you, like talking to the people your project is for, visiting partners and building relationships with funders.
Take these ideas into your own grant work
10 practical experiments inspired by the people shaping AI in nonprofit fundraising.
Get the workbook →Profile
Woodrow Rosenbaum
Main idea
Understand behavior before you optimize communication.
Who he is
Chief Data Officer, GivingTuesday
Founder & CEO, With Intent
What he’s known for
Generosity data, consumer behavior, brand strategy and using large-scale data to understand what actually moves people to act. His perspective combines sector-wide philanthropy research with a consumer-marketing lens.
Where to find him
Fundraising.AI Summit 2026
Where Giving Goes From Here: Day 1 Recap and Where The Sector is Headed
with Nathan Chappell
Why this idea matters
What makes Woodrow’s perspective interesting is that he does not look at fundraising only from inside the nonprofit sector.
His work sits at the intersection of generosity data, consumer behavior and marketing strategy. Through GivingTuesday, he works with large-scale data on how people give and participate, while his agency background brings in the question marketers have always had to ask:
What actually makes people act?
That is a useful counterweight to an AI conversation that can become very focused on production. AI makes it dramatically easier to create more content, more messages and more variations. But producing more does not automatically mean influencing behavior more effectively.
The more interesting opportunity is to use better data and better analysis to understand what people respond to, which patterns matter and where our assumptions are wrong.
What it means for your grant work
As grant professionals, we often spend a lot of time thinking about what we want to tell funders. Woodrow’s perspective suggests turning that around:
- What do we actually know about how funders make decisions?
- Which priorities appear repeatedly in the programs they launch?
- What kinds of organizations and projects have they funded before?
- Which criteria seem to matter in practice, not just in the wording of the call?
- Where are we relying on evidence, and where are we simply repeating assumptions about “what funders want”?
AI can help analyze patterns across funding guidelines, past awards, strategy documents and other available information.
The goal is not to produce a more persuasive application by guessing harder. It is to build a better understanding of the decision environment before you write.
What we’re listening for at the Summit
Woodrow brings a perspective that is quite different from most of the speakers on this list because he combines large-scale generosity data with a consumer-marketing background.
We’ll be listening for what AI changes in our ability to understand behavior, not just create more personalized communication. Can AI help nonprofits detect patterns they previously could not see? And does better behavioral insight change fundraising strategy itself, rather than simply making existing communications more efficient?
Why this idea matters
“Human in the loop” has become a standard phrase in AI. Usually, it means that a person stays involved to review outputs, make the final decision or intervene when something goes wrong.
Tim Lockie takes that idea further. In his 2026 article Humans Are The Loop, his point is that people are not simply there to supervise an AI system from the outside. They provide the intention, context, judgment and self-awareness that make the whole process meaningful in the first place.
That distinction is important. If we think of humans only as the final quality-control step, it is easy to design AI workflows where the machine does the “real work” and the person simply checks it afterwards.
Tim’s framing suggests something different:
The human role should shape the process from beginning to end.
And that connects directly to his broader work on culture and AI adoption. AI implementation is not just about giving people access to better tools. It is about building the confidence, judgment and shared understanding people need to work with those tools well.
What it means for your grant work
This shows up very quickly in grant teams. A policy might say: “Do not enter confidential or strategic information into AI tools.”
That sounds sensible. But what does “strategic information” actually mean for someone writing a funding application?
- A draft project idea?
- A budget?
- Partner information?
- An unpublished concept note?
- Evaluation findings?
- A previous application?
If people are unsure, two things tend to happen: either they avoid AI completely, or they use it quietly and inconsistently.
For grant teams, humans being “the loop” means more than reviewing the final draft. People still need to decide:
- What information should AI work from?
- What can safely go into the tool?
- Which assumptions should be challenged?
- Where is professional judgment needed?
- Who is responsible for the final decision?
That is the difference between simply adding AI to grant work and building a process where people remain genuinely in control.
What we’re listening for at the Summit
Tim’s session with Dan Kershaw is especially interesting because it combines this human-centered perspective with a real nonprofit implementation case.
We’ll be listening for what “humans are the loop” looks like in practice:
- Where do people still need to make the key decisions?
- Which parts of the process can AI genuinely take over?
- And what organizational habits make sure that human judgment does not slowly disappear as AI becomes more capable?
Profile
Meena Das
Main idea
Fix your relationship with the data before you focus on the AI.
Who she is
CEO, Namaste Data
Founder of the AI Equity Project
What she’s known for
AI equity, humanity-centered data practice, and asking harder questions about power, participation, accountability and whose data actually shapes AI-supported decisions. Her 2026 AI Equity Framework turns those concerns into a much more concrete operating model.
Where to find her
Fundraising.AI Summit 2026
Taking Your Data From Chaos to AI-Ready
with Erin Stender and Blane Warrene (Omatic)
Why this idea matters
A lot of AI conversations start with a question like: How can we use our data with AI?
Meena’s work pushes one step further back. Before asking what AI can do with your data, you need to ask whether the data itself is trustworthy, representative and being used responsibly.
That is what makes her perspective so useful. Her AI Equity work is not just about technical data quality. It also asks who is represented in the data, who had a say in how it was collected, who benefits from the analysis, and who may be affected by the decisions that follow.
So “AI-ready data” is not simply clean data.
It is data you understand well enough to know its limits.
What it means for your grant work
Grant applications are full of claims about need, target groups, reach and impact. And AI can make those claims sound extremely convincing.
That becomes dangerous when the evidence underneath them is weak. If your organization only has outdated figures, anecdotal assumptions or data collected from a very narrow part of the target group, AI can improve the wording without improving the truth of the claim.
So for grant work, the practical takeaway is simple:
Give AI trusted evidence before asking it to build an argument.
That might be evaluation reports, official statistics, policy papers, previous project data or research you have already reviewed. And if the evidence is incomplete, let AI help you spot the gaps rather than fill them with something plausible.
What we’re listening for at the Summit
Meena’s session title this year is Taking Your Data From Chaos to AI-Ready.
We’ll be listening for how she defines “AI-ready” in practice. Is it mainly about organizing and cleaning data, or does readiness also require organizations to understand where their data came from, whose experiences it represents and what it should not be used to infer?
Steal Their Best Ideas10 practical experiments inspired by the people shaping AI in nonprofit fundraising.
Get the workbook →Why this idea matters
Dan’s work stands out because he is treating Furniture Bank almost like a live case study for nonprofit AI adoption.
Instead of waiting until everything is perfectly figured out, he documents the decisions as they happen: what they are testing, what the team is ready for, what still needs more guidance, and where the answer is simply not yet.
That is where the idea of an AI adoption playbook becomes interesting.
The playbook is not something written first and implemented afterwards. It emerges from real use.
Teams try things, notice where people get stuck, identify what guidance is actually needed, and then codify those lessons into more repeatable ways of working.
That makes his work a useful counterpoint to AI adoption strategies that start with a long policy document and only later ask how people are actually supposed to use the technology.
What it means for your grant work
Grant teams can take the same approach. You do not need to design the perfect AI system before anyone starts using it.
Start with one real workflow, such as funding research, developing a project idea or reviewing an application. Document what works, where people get confused, what needs human judgment and which instructions are worth keeping.
Over time, those experiments become your own practical AI playbook for grant work.
What we’re listening for at the Summit
Dan’s session with Tim Lockie should give us one of the most concrete implementation perspectives of the Summit.
We’ll be listening for what Furniture Bank has learned from building its approach in real time, which practices have become repeatable enough to codify, and where Dan still thinks the most responsible answer is “not yet.”
Profile
Cherian Koshy
Main idea
Donors are not purely rational decision-makers, and our fundraising strategies should stop pretending they are.
Who he is
EVP, Kindsight
Author of Neurogiving: The Science of Donor Decision-Making
ACFRE
What he’s known for
Donor decision-making, behavioral science, and challenging fundraising advice that is based more on assumptions than evidence. His work looks at how memory, emotion, identity, trust and timing shape giving decisions, often outside conscious awareness.
Where to find him
Fundraising.AI Summit 2026
Donor Behavior Is Changing: How Leaders Must Adapt
Why this idea matters
A lot of fundraising practice is still built around a fairly rational model of the donor. We assume people carefully consider the information we give them, weigh up the options and then make a decision.
Cherian’s work starts from a different place. Giving decisions are shaped by memory, emotion, identity, trust and timing. That does not make donors irrational. It makes them human.
What we especially like about his approach is that he is also very cautious about overstating the evidence. Not every behavioral study applies to every fundraising context. Some research is thin, contested or based on populations that may look very different from your donors.
So the lesson is not simply “use more psychology”.
Understand how people actually make decisions, and be honest about what the evidence can and cannot tell you.
What it means for your grant work
Grant professionals are already making a lot of assumptions about how AI is changing funders:
- That they are receiving far more applications.
- That proposal quality is becoming more superficial or generic.
- That funding is getting tighter.
- That personal relationships will matter more.
- That some funders may respond by closing open calls or moving toward longer-term partnerships.
Some of those things may well be happening. But we do not yet have enough evidence to treat them as facts.
Cherian’s perspective is a useful reminder to separate what we know from what we assume, and to look much more closely at how funders are actually changing their behavior before redesigning our own strategy around those assumptions.
What we’re listening for at the Summit
Cherian’s session focuses on changing donor behavior.
We’ll be listening especially for what he is seeing around trust: how AI-generated content and more automated fundraising affect donor relationships, whether similar trust shifts are showing up in the US, and what nonprofits can do to avoid eroding trust as AI use grows.
What we’re taking into the Summit
One of the reasons we love the Fundraising.AI Summit so much is that it kind of resets our brain once a year.
Because if you spend enough time in general AI content, especially on YouTube, you quickly end up with some 23-year-old guy telling you that this one new AI agent will replace half your team by Friday.
And that is just not the conversation we want to have.
The Summit has always felt different. From the beginning, there has been a much stronger focus on responsible AI, trust, values, ethics and what all of this actually means in the nonprofit sector.
Not just: What can AI do now?
But also: Should we do it? Does it actually help? What does it change for people? Does it create positive societal impact? And how do we use it without losing the things that actually matter in our work?
That is why these two days feel so valuable to us.
We’ll keep adding to these profiles as we go and see which ideas hold up, which ones change, and which ones we want to take back into our own grant work.
Afterwards, we’ll turn the strongest ideas into the post-Summit edition of our workbook.
Steal Their Best Ideas
10 practical experiments inspired by the people shaping AI in nonprofit fundraising.
Get the workbook →