Fundraising.AI to Lead the Nonprofit Sector With Artificial Intelligence Best Practices
Nathan Chappell wanted to create guardrails surrounding the use of artificial intelligence (AI) in fundraising four years ago, but there wasn’t an appetite for it yet.
It took a broader interest among his nonprofit sector peers — and the release of ChatGPT last year — for that vision to come to life.
Through those early years, small groups would get together to discuss the possibilities on chat forums, in Zoom meetings and at in-person nonprofit sector conferences, but nothing really came of the informal discussions until after a breakfast meeting at April’s AFP Icon in New Orleans.
Afterward, Chappell reached out to 125 subject-matter experts — 92 of whom were eager to help drive the effort forward. And 1,700 emails later, the group created what became Fundraising.AI’s Framework Toward Responsible AI for Fundraising that was released in July — just three months after the effort began generating steam.
Chappell bought the fundraising.ai domain a decade ago with no specific intention for it. He now uses that site to encourage individuals and organizations across the nonprofit sector to endorse the framework’s 10 tenets, or best practices for fundraising with AI, that revolve around:
- Privacy and security
- Data ethics
- Inclusiveness
- Accountability
- Transparency and explainability
- Continuous learning
- Collaboration
- Legal compliance
- Social impact
- Sustainability
What the Framework Means for the Nonprofit Sector
When Chappell, who also serves as senior vice president at DonorSearch AI, first considered the possibility of the framework, he combed through countless AI frameworks — some of which were created for the social sector. Still, he felt a framework specifically for fundraising was needed.
“I noticed that while the word ‘trust’ was being used widely in other frameworks, that ‘trust’ was referring to AI essentially not harming humans,” he said. “And when we think about trust in the context of fundraising, it's much more intimate than that.
“It's essentially the basis that nonprofits operate, from a fundraising perspective. So that led me to this idea that if we don't create our own framework, one will be created for us. And so that was really the moment that I was like, we have to, as a nonprofit sector, take charge.”
Chappell admitted the framework isn’t perfect, but that’s not the point. Instead, it achieved the consensus of 92 thought leaders in the nonprofit technology space. A 30-person advisory board will now review it on a quarterly basis to ensure it maintains the best practices as the technology evolves.
“A great example is, if we had built the framework before November of last year, we would have not really included generative AI,” Chappell said. “And now, when we built it, we really had to think through, how does this apply to generative AI? And there's still many more technological kinds of changes that we're gonna have to [accommodate].”
Nonprofit Technology Vendors Join Forces for the Sector’s Benefit
One of the tenets of the framework speaks to collaboration. Specifically, those who endorse the framework, agree to “actively engage with and learn from my peers in Fundraising.AI, sharing my experiences, challenges, and successes in Responsible AI for fundraising.”
Chappell described the collaboration to date as remarkable, but it’s also part of how those in the sector think. There was a fear among those involved that this could instead become a for-profit initiative if they didn’t act quickly.
Dawn Galasso, with more than 10 years of experience in the nonprofit technology space, has worked for other nonprofit tech vendors prior to joining GivingDNA as vice president of technology sales in 2021. Therefore, she knows many people at competitor fundraising technology software providers, but she and her peers are able to set aside that healthy competition for the good of the sector.
“This, in particular, has brought us together because it’s a common concern that’s new, and we all love our nonprofits,” she said. “We all are trying to help them do more good in the world. That’s our No. 1 reason. We can still compete with each other, but, as a team, making sure that our industry stays safe and stays strong allows us to continue to compete with each other in the right way and not have disruptions.”
Nonprofit Technology Vendors React to the Framework
The framework separates companies willing to slow down for the benefit of the sector versus those not willing to lose a potential market advantage, but endorsing the framework is an easy call for many vendors. Though AI is a key part of Neon One’s strategy, the software provider will only develop technology through the lens of the framework now — even if that requires it to move slower.
“Because there's some companies that might not adopt the framework because they feel they will lose a market advantage by not being able to move quickly,” Tim Sarrantonio, director of corporate brand at Neon One, said. “And what happens in the technology sector for nonprofits is that there's not enough listening, and there's too many assumptions made on what's good for nonprofits.”
Meanwhile, with data privacy regulation changes and a targeted discussion on how data should shape AI already happening internally, GivingDNA is among the early endorsers that sees the value in addressing AI hot topics, like how to avoid biases in AI models.
“An example would be we may have a bias around people in different generations — what they’re interested in, what they’re not interested in, how are they responsive?” Galasso, one of the 92 thought leaders involved in shaping the original framework, said. “If we incorporate those biases into the AI we’re building to be able to segment an audience much more quickly, potentially that could be detrimental.”
Sarrantonio — a Fundraising.AI founding member alongside Chappell; Mallory Erickson, creator of the Power Partners Formula; and Meena Das, CEO of NamasteData — added another concern he’s seen in the nonprofit technology space — an AI wealth screening tactic that upsells donors solely for living in affluent zip codes.
“That's bad,” he said. “That's a toxic generosity experience for people. And what we do is we want to celebrate healthy generosity experiences. So only if it builds trust between the fundraiser and the donor — that's what we're going to focus on.”
What the Framework Means for Nonprofits
The initiative has received about 150 endorsements at press time, though the true number is likely much higher with many nonprofit technology vendors telling their entire staffs to abide by the framework. Though a lot of endorsements have been from vendors, nonprofits and agencies are also in the mix. And there are seats at the table for nonprofits as well.
“It’s not just vendors saying, ‘Oh well, we’re embedding it into our platforms and our products,’ and ‘Are we doing it ethically?’ but it’s also nonprofits weighing in on what they think is ethical also,” Galasso said. “So, it’s a good balance of people in the room, really good thought leaders in the room having these conversations early in the game.”
It Can Help Nonprofits to Vet Vendors
Nonprofits should question their vendors to ensure a good fit, but may struggle to guarantee their vendors meet ethical requirements. Going forward, nonprofits can rely on the framework to help them ask the right questions when buying an AI solution. After all, in the social sector, nonprofit vendors should be transparent with their nonprofit clients.
“The framework itself for a nonprofit actually serves as like a vendor questionnaire because, and our hope is — and this is our more esoteric hope — that Fundraising.AI becomes like the Good Housekeeping Seal of Approval for responsible AI,” Chappell said. “And so meaning that, if a nonprofit determines that they are ready to buy AI, that they would only buy it from a vendor that has signed or endorsed the framework.”
To help nonprofits even further, vendors that have endorsed the framework will soon be able to display a logo of that seal of approval on their websites.
It Can Help Build Trust
Galasso views the framework as a handshake between vendors and nonprofits, and, in turn, between nonprofits and donors. That way, nonprofits can be assured their data is in good hands when they share that data with their software providers. Having this ethical standard in place could prevent disaster for the sector.
“If you had one vendor do something that’s misplaced, and a couple of nonprofits were affected by that, you know that that’s a news story,” she said. “We have to put those rails in place ahead of time where we can so that a nonprofit can feel confident in the vendors that they’re choosing.”
AI Facilitates Your Work, Does Not Replace You
This framework is made for nonprofits of all sizes, so Sarrantonio has made specific efforts to include small and mid-sized nonprofits with revenue less than $5 million. Those organizations, he said, don’t know where to start when it comes to AI and often have imposter syndrome.
“And that's what's different about this is that their inclusion is part of the whole plan and part of the whole growth of this,” he said. “That's where my obsession is personally going to be as well. But they're worried that this stuff is going to put them out of work. And I have to go, ‘No, it's going to accelerate your ability to make an impact.”
Galasso agreed the fear is misplaced.
“There still needs to be a person that’s doing that end job,” she said. “So, what I would suggest to nonprofits is that by embracing it, it’s actually going to allow them to do a better job at fundraising by taking away those mundane tasks that weren’t really getting them to goal anyway. It was all that extra work and now they can do more things that they should be doing — raising money, going out and talking to people.”
What’s Next for Fundraising.AI?
Aside from the framework, the team of almost entirely volunteers has been busy planning its inaugural event, the 2023 Fundraising.AI Virtual Global Summit. The conference, which takes place Oct. 23-24, is free to attend with a track for nonprofit practitioners, as well as another for nonprofit platforms.
The nonprofit track will tackle issues such as AI ethics, AI’s potential impact on donor trust and its role in nonprofit professionals’ jobs, while the vendor track will share insights into consequences of its irresponsible use, how to empower nonprofits and ways to protect donor data. Organizers have reached its goal of 3,000 attendees already and have raised more than $250,000 plus in-kind support from some of the biggest technology companies in the sector, including AWS, Blackbaud, Salesforce and Microsoft.
Though Fundraising.AI is not currently a nonprofit, Chappell didn’t rule that out as a possibility in the future. For the time being, GivingTuesday is supporting the initiative as Fundraising.AI’s fiscal sponsor since the initiative has not yet become a legal entity. Essentially, that allows nonprofit vendors to support the effort and sponsor the event by writing checks to GivingTuesday, which, in turn, designates the money to the Fundraising.AI initiative without taking a cut of the money.
Despite that setup, Fundraising.AI isn’t a small initiative.
“Fundraising.AI represents one of the next evolutions in governance in our sector, where ownership is a collective goal, an individual company's goal. And it is something that is owned by the sector itself in many ways — and that's important, but you have to evangelize that. You have to be very focused on protecting that.”
In fact, it’s gaining international attention, with Chappell becoming the face of the movement in the nonprofit space by serving as a member of the artificial intelligence committee for the InterNational Committee for Information Technology Standards (INCITS).
“Our work has just started,” Chappell said. “And I think [members of] this body will be the ones that are really creating those best practices worldwide, which I think was unexpected, but really, to me, a really exciting part of what we're doing.”