Nonprofits and AI: 92% Are Using It, But Only 7% Use It Well

Hey, welcome back everybody. It's another episode of the nonprofit show, but not any episode of the nonprofit show because it's nonprofit power week and we are with our friends all week long, every day from with our friends from YPTC, your part-time controller. Today we are welcoming Patricia Buessel. She's a client services technology adoption manager. That's a that's a mouthful, isn't it? Welcome, Patricia. >> Thank you, Julia. Thank you for having me. >> It's going to be a lot of fun. We're going to be talking about adopting AI that you can actually trust. And that is such a big and scary thing. Um, and so, Patricia's going to help us understand what we need to be thinking about and what we should be looking for. Um, we get to have these great conversations in Power Week because we have nonprofit sponsors that are with us day in and day out. They include Bloomerang, American Nonprofit Academy, Staffing Boutique, JT Consulting, Third Sector Company, Your Part-Time Controller, and Martis. They are a new sponsor, and we'll be talking more about Martis um shortly. We're very, very excited to welcome them. Welcome them. I'm Julia C. Patrick, CEO of the American Nonprofit Academy and Patricia Weso, client services technology adoption manager. as I mentioned over at YPTC, your part-time controller. Okay, I got to believe that your phone is ringing off the hook because this is a big topic in the nonprofit sector, technology, and how do we adopt it? Tell us what it is you do, Patricia. >> Thank you, Julia. Yes. So, uh this is a reciprocal we just started this year at YPTC where uh my aim is to thoughtfully, you know, evaluate, adopt and scale technology um whether it's AI enabled or just regular technology across um our client service delivery because what we want to do is empower our staff and empower our clients to really use technology thoughtfully. Wow. It's such an interesting time because I've got to believe what you are advising your clients on, it's like changing every 10 hours, right? I mean, how how frustrating or liberating maybe is it that you get to be at the center of this discussion? >> Yes. Well, it is, you know, but you um what we like to think about, we have a a director of AI here at YPTC as well, and what the base doesn't really change, right? So, the right tools, the governance, um in the end, AI is a software and a tool like a lot of the other technology tools that are out there, right? So we have to really kind of lay that out there as we evaluate and guide our staff and our clients on adopting it. So >> yeah, it's interesting and I think that's a healthy way to look at it, right? Because then you're not coming to it from a point of fear because if that's if you're fearful when you go into these discussions, that's just, you know, as we say in my family, no bueno. It's not a good way to go. Well, let's dig into this because we have a lot of questions to ask you, a lot of things um to to talk about. But this is one of the things that I'm the most interested in today is experimenting to real capability. Meaning what? Because experimenting and AI o that seems a little frightening. Well, you know, there were um there was a recent survey um from virtuous uh AI. It's a software and the fundraising AI um that talked about how 92% of nonprofits are already using or experimenting with AI. Um we did a recent survey internally as well and it's it's very similar, right? Um we maybe around 10% of our nonprofit clients are not using AI. So that means that everyone is it's not really an adoption problem. People are using it. They're experimenting with it. Um so it's important to start seeing how we can maybe take this at scale and start realizing the impact we wanted to have because even though 92% of nonprofits are using it, only 7% are using it in a meaningful and impactful ways. is what the survey um mentioned, right? So, that's really um interesting to start evaluating. So, so how are these nonprofits using it? How are they experimenting? And that's a great point you brought in um earlier. And what I want to just kind of talk about is that there are very um simple ways to get started. It can be very daunting. Um but there are simple ways like we have we have used these and and some ways to start thinking about it is maybe start making a list working with each of your departments and start working start making a list of use cases and workflows that can be potentially supported by AI. So having that in place can really kind of start um giving you a framework of where to get started, right? And then as you're working on that list, rate it, right? Rate it. What is the level of effort you you currently h have? What is the value this workflow or use case brings? And then finally, what is the risk that it involves? Right? Um so kind of breaking it down and by workflows or use cases. um really kind of helps you just get started, start laying out the foundation of where to focus on, right? Because I know AI can be daunting. Um >> so >> I like that because I think that's a really smart way to be thinking about what it is you need to be doing, like what's your workflow anyway? Where where are things, you know, bottlenecked and where are things moving quickly? Um, I think that's a really smart smart way to go because then from there it seems to me like then you can be a lot more intelligent about what risk you have. So I want to talk about that because you're advising us to really think about this concept that not all AI carries the same risk when you use it. And if I could back up a little bit, talk to us about what risk is when we are talking about AI. I think a lot of us in the nonprofit sector, you know, we understand risk. We think about it in a human centric way like how are are we keeping our employees and our clients safe, but what is this new this risk factor that we should be thinking about? So yeah, a good way to start assessing risk and evaluating risk is back to the workflows and use cases. Um we talked about rating it right as going through each of them and rating it. So ask yourself, right? What happens if the AI is wrong? What will be the impact if we use AI for something and it it gives us a wrong answer? It is biased, it is insecure or we don't use it properly. Right? So if you um scope it down to a workflow level, it's easier to start analyzing it and and assessing it and then you can start addressing it, right? Um so we we talk about maybe rating it into um low risk, medium risk, high risk, right? Think about lowrisk cases as minor inconveniences, right? You can use it for brainstorming, rewriting emails, maybe generating outlines, uh summarizing meeting or presentation, uh notes, right? Um those are all things that are low risk and probably it's potentially what most people are using it for, right? Um and then we go to medium risk. So think about medium risk as um this is where AI can influence decisions. of course, it's not making the decisions for you, but it can help it can influence um what you're looking at. Um so maybe when you use AI to analyze financials or you use AI to segment donors into different categories, right, it's not necessarily giving you an answer, but it is influencing you into looking at things a certain way, right? >> Yeah. And then finally the high-risk cases, right? Which I believe like um I don't know if many people are using it for high-risisk. I don't I'm not sure. I don't have any data on that. But um high-risisk cases are those where um it makes or can heavily influence a decision for you, right? And we don't really want that to happen right now. um it can have impact on people on money um or your or the organization's reputation, right? So um all of these so these are the three levels of risk like so so just think about it as as think of of this as when you're evaluating your workflows for use cases um assess what level of risk if I add AI to this what level of risk will it involve right and then that will then say drive you to ask yourself so where should human judgment and controls exist through red we >> yeah I love that you layered this in and I really appreciate that uh you've painted a picture as to what risk factors are because risk is different and the way you just laid it out was really smart um because it's not that it's not that overwhelming to understand >> right it's I almost want to say it's pretty basic and clearcut cut in some forms uh of looking at this. And so we understand that you mentioned something and I want to talk about this that this human judgment is the most important control. Um how do we get that translated and in practice across our organizations because it seems to me that um we're not doing that. We're just saying, "Oh, great." You know, we'll take what what it spits out, right? >> But not not so fast. >> Yeah. So, so that's why it's understanding and assessing risk is easier when you break it down to the workflow level and not try to address it as a whole. I mean eventually you will have to address it as a whole but um you can get a better sense of it at at at each of the workflow levels right so um the goal is not to really eliminate the risk but manage it and how so how do you manage it by incorporating human judgment into the steps right >> right I like that you said that because that changes the thought process right? You're going to have risk. You're not going to get rid of it. It's just how do you manage it? And I love that concept of using that word judgment. Um because you know AI is does make mistakes. We make mistakes. Um we know the power of the prompt. The way you prompt AI um completely changes how what your results are. I think that is the untapped brilliance of AI is understanding that prompt and I'm gonna use the word relationship. Um how do you navigate this? Um and again it's human judgment. It's very yeah it's it's very interesting to me. Really really interesting. Okay. So you've talked to us about risk. You've talked to us about human judgment. I love the understanding the workflow process. We we layer all these things in the next big thing and to me this is the big topic of the day is where does your data go and do you even know where it's going? Talk to us about this because this is somewhat like I don't think we can answer this question. >> Yes. Yeah. Um, so having the right tools in place is truly important to be able to scale AI across your organization and where you add your in like so the tools are what define or determine where your data goes, right? Um, at YPTC we don't like we don't like to recommend a specific tool. Uh but we do like to si tell our nonprofit clients that um usually the safest place to start is to use the built-in tools that already exist within your productivity suites, right? So, for example, if the if your nonprofit uses Microsoft Office or if you use Google Gemini, then um using Microsoft Copilot is probably the best route for you because um because the tools already exist with the rest of your data and productivity tools. So, it's easier to manage where your data goes and and to control it and configure it appropriately. >> Yeah, I I really noticed. In fact, I was talking to our executive producer, Kevin Pace, um, about I'm getting to the point where if I'm using some sort of platform that we use here for our business, um, and there's not a an internal AI feature, I'm kind of bent out of shape because it's so easy just to be like, oh, hey, change this or how do I do this or where do I find that? And um I have the sense that if I am working within that software that it's more safe. I mean is and you just mentioned I mean is that that's fair to to think >> it's important to always go in and review and configure make sure the tool is configured correctly. if it has an AI um embedded in it, it's always important to go in and and make sure that it's configured appropriately no matter what tool, right? Whether it's one that's already built in or it's a separate standalone tool, right? So, um you really have to watch out for that. And also try to use tools like when when you're selecting a tool, try to use ones that do have suck to guidelines. I mean some nonprofits um GDPR compliance have some what of a HIPPA if if your organization deals with HIPPA um are not have an understanding of what that is right as well. Um but mainly what we like to watch out for is for sock two or sock 2 compliance. So um keep that in mind as you are evaluating your um your AI tools and configuring them as well. So >> So let me ask you about this because um a little bit more >> deeper I I think more deeply. It's um so many of us in the nonprofit sector are really enjoying the advantages of having fundraising software that's on a platform that really is powerful and we're uploading so much sensitive information from credit cards to relationship management notes, you know, legal documents, especially if we're dealing with donors that are working with estates or larger gifts. How do we navigate the risk mitigation and the safety of the data? Because we get I think we get like really charmed by oh my gosh look what this will do and then we're not thinking okay but where does it live? I'm how do we go about that process of even asking the question? >> That's a great question. And it's always very important when you're as you're evaluating the tools that that you choose, right? Um like you said, um you can very be very dazzled by a really AI forward agentic AI tool that is going to solve all of your problems. >> Um but it's always important to stay grounded. Again, it's a tool they need to have. Um, you need to ask about where the data goes, how we can configure it AC appropriately and work with IT, if you have an IT security team or maybe perhaps um a board committee um that does understand AI and technology that can help you kind of ask these questions and configure the tools once you've chosen them. Right? So again, don't be afraid to ask the questions um and and review and configure it before you start using it. I know it can be very um enticing to just go in and and get started, but it's always important to watch out for where the where how what they do with the data, where do they store it, um do they how long do they keep it, right? what happens if you >> terminate the rel terminate the subscription or something what will they do with the data once >> um it's important to ask these questions as you're evaluating the tools >> right I've got to believe too that understanding like do they carry insurance or how do they protect you or even down to like how quickly do they notify you what's their protocol if there's a breach um is is that fair to ask those questions. >> It is fair. Yes. Yeah. Very fair. Um >> what you're right. That's exactly what is there their security protocol. Um that's very important. >> Yeah. I've got to believe that we haven't and across the the landscape of the nonprofits, we probably don't have protocols for that ourselves, right? Um, but we sure should know what our partners and our our platforms are doing because that's it's just so scary to think about that. Um, and I think it's really frightening not to know the answer because I feel and I don't know if you're seeing this, but I feel like donors are now asking those questions because they're hearing about it. you know, just getting ready for work this morning. Um, on my local National Public Radio station that I listen to every morning as I'm getting ready to leave for the office, there was a huge article about AI um impacting leaving its its, you know, home base and doing things it shouldn't have been doing, right? And so that's a discussion that is chilling, but we don't know what we don't know, right? >> Yes. Yes, >> go ahead. >> That's right. No, and and that's why it's also very important to um start um if you want to if you want to again focus if your skill is on if your if your focus is on scaling AI and technology with all of this all of the things that we're hearing about and um are going on and we feel like we have no control over it's important to focus on what we can control and that is on establishing guidelines and ensuring that maybe an acceptable software use policy is in place so that people know your staff know what is encouraged, what needs approval, and what is off limits, right? Because all of this um can be very daunting. >> Yeah. Yeah. Absolutely. It's fascinating um to me to kind of step back your advice about um evaluating things a lot more. I didn't expect for you to to lean into that, I guess. But that's absolutely the smart way to go. I totally see where you're going with that and that mentality ultimately can help you and protect the organization. Um, and so that leads me to my last point that I want to ask you about. And you're advising us that governance should happen before technology. What does that mean exactly? So, as I was mentioning a bit ear um before, it's important to kind of be able to to have a list um simple guidelines in place that can um allow your staff to confidently use AI and technology tools. Right. At YPTC, we like to say um that guidelines help protect the adventurous and reassure the cautious. >> I love that. That's a great great line. Per That's so true. >> It is. It is because we want we want everyone to also the idea is not to um stay away from the technology. It is to embrace it. But the way we embrace it is by maybe building some guard rails around it, right? Um so so that's why we like to say governance before technology. But um I know it's also something that can be seem overwhelming for people and for organizations like where do we get started, right? Um but again think of AI as another software, another tool. So, uh, perhaps your organization already has an acceptable use policy. So, just incorporate AI into that software or acceptable use policy. Um, so that's just a list of approved. So, you can add a list of approved tools so that staff know um what they're allowed to use with your uh organization's data, right? >> And then um and then start working on guidelines. um make sure that they align to your mission and values. So consider what you and your organization is comfort comfortable with, right? And what the vision is, what the mission is. So um get started with guidelines. Um make sure you you communicate these and and then ask for feedback. Right? So again, we'd like to be these guidelines can start simple, but be clear on what is encouraged, what needs approval, and what is really offlimits, right? And then ask for feedback. Um, so just get some guidelines going. I know a lot of we've talked about tools um we've talked about tools such as Microsoft and Google and fundraising platforms. A lot of them have frameworks available that can help you get started. I know uh both Microsoft and Google Gemini have um special offers and frameworks for nonprofits that if you don't know where to start perhaps um researching those can help can help you get started. Um and then also uh an important piece of governance is training. So you talked about right what how do we prompt it? Um how do so just training staff is also important. I think that's one of the biggest hurdles of of scaling the technology is like where do we get training? Um but again look for there's there are a lot of resources out there. Um I I I don't want to say it again, but Microsoft and Google, if you are using those tools, they offer uh pre-learning paths for uh for staff to to be able to know how to use the tools, be able to prompt them. So, take advantage of those resources that are out there. >> Right. I am amazed. I mean, you brought up training. Um I mean, I'm obviously with the American Nonprofit Academy. We do a lot of training around the world and I am amazed at how much training there is now on AI from all different parts of the of the business world. Um, and a lot of it is training on use. >> You know, how do you use this technology? Um, it's not fear-based. It's not, you know, what not to do, but it's more thoughtful like what can it do? How do you manage it? How do you manage expectations? How do you keep yourself safe? your organization. Every day I'm just seeing more and more of this type of content coming in uh to our company. And I think it's a it's a really important um aspect of managing uh this technology um because as you said right right in the beginning, it's the human judgment that we need to lean into, you know, because we're having AI work for us. We shouldn't be working for AI in essence. Right. >> Exactly. AI is here to support us and support our judgment, not to make the decisions for us. Right. So, >> well, I I love it. I think it's um a huge gamecher um in our business and in our sector. I just think it's it's an amazing technology, but it's not without its risks. And and to your point, you have to be thoughtful. And I love that you used that word govern. You know, we have to look at governance of it. And uh yeah, super cool. Well, this has been a great conversation. I feel Patricia Weso that if we had you back in even three months, so much is going to have already changed, right? This is a moving story and a moving target. And this is why we've had Patricia um on today. She's a client services technology adoption manager over at uh YPTC, your part-time controller. You can learn more about the work that YPTC does at ytc.com. Um it's not just about having um bookkeepers on your team anymore. You need thought leaders. You need practitioners of all these different types of things that YPTC is really leaning into. and Patricia is a perfect example of that. And so this has been a fun conversation. Thank you for joining us. >> Thank you for having me, Julia. And yes, I'll be here in in three months if you >> That's right. You'll be here and you'll have a lot more to say, I'm sure, because things will change. I might hold you to that. You know, Patricia has joined us um today on nonprofit power week. And as I mentioned when we kicked off the show, we don't do this very often, only a couple times of of the year. So, this week we're talking about, as we did today, adopting AI you can trust. Yesterday, we had an amazing conversation about red flags that boards miss and that they need to know about. We're going to be talking about what do you do if you need to change your accounting partner and the questions that leaders should ask their finance team. Um and then before you buy, what are five questions about software? Um and so a lot of different topics, but they all weave together um during nonprofit power week. And so we are delighted that you could join us. We're also delighted that we have amazing support and they come from uh our partners at Bloomerang, American Nonprofit Academy, Staffing Boutique, JMT Consulting, Third Sector Company, of course, your part-time controller, and our newest sponsor, Martis. We are so delighted that these folks step up and join us. As we leave each and every episode of the nonprofit show, we leave with this message and it goes like this. to stay well so you can do well. We'll see you again.