Responsible AI Adoption & Ethically Using AI

Tony Martignetti
Hello and welcome to Tony Martignetti Nonprofit Radio, big nonprofit ideas for the other 95%. I'm your aptly named host, and I'm the podfather of your favorite hebdomadal podcast. Oh, I'm glad you're with us. I'd be forced to endure the pain of endo-ortitis if you inflamed me with the idea that you missed this week's show. Here's our associate producer Kate with What's Up.

Kate Martignetti
Hey, Tony, I'm on it. Responsible AI Adoption. We continue our coverage of the 2026 Nonprofit Technology Conference with a panel that helps you find the low-hanging fruit for AI at your nonprofit. They share their 5-step framework for deploying AI in small, thoughtful steps. They are Jen Frazier from Firefly Partners, Kate Dreyfuss at New Music USA, and Ben Freda with BFC Digital. Then Ethically Using AI. Chris Rosica and Bethany Friedlander share 5 strategies to create compelling web and social content with AI, keeping in sight ethics and risks: ideation, targeting, optimization, repurposing, and staying human. Chris is from Rosica Communications, And Bethany is at New Bridge Cleveland. On Tony's Take 2.

Tony Martignetti
Tales from the Gym. Meet Chatty Sam.

Kate Martignetti
Here is Responsible AI Adoption.

Tony Martignetti
Welcome back to Tony Martignetti Nonprofit Radio coverage of 26NTC. You know that that's the 2026 Nonprofit Technology Conference. That all these smart tech folks and tech users— this is not strictly a conference for tech employees, tech professionals. This is also for all professionals who use technology. So unless you're still using index cards as a CRM, you are using technology, odds are. You're using Microsoft 360, most likely. So 365. 365 or 360?

Jen Frazier
365.

Tony Martignetti
365.

Jen Frazier
It's all— it's every day of the year, Tony.

Tony Martignetti
Every day. Well, it could have been 360 degrees. I did take 5 days off, all right? It could have been 360 degrees, degrees of a compass. I was a Boy Scout. Now you're mocking the Boy Scouts?

Jen Frazier
I am.

Tony Martignetti
We haven't even introduced— all right, I'm shutting our mic off. We now have 2 panelists instead of 3. All right, so we're at the 2026 Nonprofit Technology Conference in Detroit at Huntington Place. With me now were going to be Jen Frazier. She's no longer, but she's still She's no longer a guest, but she's still CEO and founder of Firefly Partners. Also Kate Dreyfuss, Director of Communications at New Music USA, and Ben Freda, President at BFC Digital. Almost like BFD. Should have been BFD Digital. Big fucking deal.

Jen Frazier
That's right, BFC, man.

Tony Martignetti
Whatever.

Ben Freda
There's still time to change it.

Jen Frazier
I mean, he changes that stuff all the time, so.

Tony Martignetti
All right. And their topic is Low-Hanging Fruit: Clearing a Path for Responsible AI Adoption at Your Organization. I trimmed off at your organization. I think that's understood.

Jen Frazier
Yes.

Tony Martignetti
But Low-Hanging Fruit: Clearing a Path for Responsible AI Adoption. Jen, you're the best person, I think, to give us just an overview of the topic. Please. Please.

Jen Frazier
Yeah. So I think right now, obviously, lots of folks are excited and a little bit overwhelmed by sort of the idea of AI and what can I do with this amazing kind of new technology that has come roaring into our lives. And so what we wanna do is just give a quick overview for folks about how to take some of the overwhelm and to really just boil it down to some pretty concrete, simple steps to get started. So we have a nice 5-step framework that— Yeah. Ben and I, well, Ben really developed, and Ben and I have been using with organizations to sort of help them just really take all these ideas and all this sort of hype, all the crazy, and boil it down and get simple and get clear with inside your organization and take some concrete steps. Because there's so much that is possible. And so we really just wanna make sure that people are like, how do I get out of like, you know, I'm paralyzed with the overwhelm and get out of that space and really start taking some steps forward. Yeah.

Tony Martignetti
All right, thank you. We're going to turn to Ben then since we have a 5-step framework, we've got a step-by-step roadmap, and then Kate will figure out— we'll see how you— so it looks like 2 consultants and a client. Is that essentially the— okay, okay. The client is flanked. She's a bit—

Jen Frazier
She's flanked. She can't get out. She's the center of the universe.

Tony Martignetti
Kate is sitting in the middle between Jen and Ben. Oh, Jen and Ben. Jen and Ben. Look at that. It all works out. All right. All right. So, Ben, why don't you want to acquaint us? Why don't you just like tick off the 5 steps and then we have time to go into, but just tick off our 5-step framework.

Ben Freda
Totally.

Tony Martignetti
For AI adoption. That is low-hanging fruit.

Ben Freda
The low-hanging fruit. That's what we're focused on. So we're, you know, so many of our clients ask us, how do we start? Right? Because they've heard all about this AI stuff. How do you start? And so we're trying to give people a real simple sort of process to follow. Just the simple stuff. We're not asking people to do big projects. We just want them to dive in to start to feel like they can get some workflow improvements, some efficiencies. So there's really 5 things you got to do. Got to do them real quick. Let me see if I can remember them. I know. If I can't, somebody will fill in. Kate has done this in her real organization for— in real life. So rather than it being all theoretical, she can give us the real down low. But

Tony Martignetti
that noise is Ben slamming his hand on the table for emphasis. We have to admonish Ben. Stop doing that.

Ben Freda
I like to add a lot of extra junk noise.

Jen Frazier
I'm usually doing this that's quiet.

Tony Martignetti
Yeah, she shakes her hands. I just shake my hands. That's fine.

Ben Freda
That's a lot of fuzz.

Tony Martignetti
Okay, Ben, go ahead.

Ben Freda
Okay, 5 steps. Ready?

Jen Frazier
Yes.

Ben Freda
So I can do it. Number 1, learn. Spend 2 hours getting all your staff up to speed on the basics of what AI is, what it's good for, what it's not good for. Okay. Number 2, jam sessions. Just brainstorm with people about their own jobs, not about AI, but about their own jobs. What do they hate doing? What's annoying? What makes them groan when they think they have to do it, right? Step 3, toss every idea you've heard into a big list. So step 3, big list. Step 4, prioritize, right? So look at everything on that list, figure out what's high effort, low impact versus high impact, low effort. Prioritize according to that. Step 5. Step 5, pick your top 2, 3 things and write up a 1-page description of it. Blueprint. What is it good for? Why are you doing it? What might the technology be involved— what technology might be involved in doing it? Then you run with the top 2, 3. That's it.

Jen Frazier
Run.

Tony Martignetti
Okay. All right. We're going to turn to Kate. What was New Music USA facing before the innovation that That BFD— no, I was not— Go with it.

Ben Freda
I love it. I love it.

Tony Martignetti
That BFC and Firefly brought to you, to New Music?

Kate Dreyfuss
Sure. So New Music USA, we are a national nonprofit organization that focuses primarily on grantmaking for musicians and arts organizations. And we're a team of 10 people.

Tony Martignetti
Do you do grantmaking to individuals?

Kate Dreyfuss
Yes, individuals. Ensembles, venues, and we also provide mentorship opportunities as well. So where we were, I think some of our team was using tools like ChatGPT on their own, but without any guidelines or real understanding of how to use it. And our connection with BFC is they helped us build our website about 5 years ago.

Tony Martignetti
BFD.

Ben Freda
BFD, yeah. As of now.

Kate Dreyfuss
As of now.

Jen Frazier
It's over.

Kate Dreyfuss
But I've been working with Ben and his team since I joined the organization in 2023. And Ben and I were just chatting about AI in general, and he offered a training for our organization back in the fall of 2024. So, we all got together and just had a learning session with Ben about what large language models are and how they work. And I think in the beginning, we were all quite intimidated, and we just didn't know where to start. We know that these tools are important and that everybody is going to be using them, and everyone is starting to use them now. But it's, it's hard to know where to start, especially if you are, you know, a team of 10 with lots of competing priorities and tasks. It can feel overwhelming. So that's, that's where we started. Just— Okay. Yeah.

Tony Martignetti
Let's start with, uh, with the first step in the framework. Let's stick with you, the, the brainstorming. Uh, how did everybody, everybody Did everybody contribute? Was it—

Kate Dreyfuss
So, so Ben and I worked together to basically decide which members of our team would make the most sense to brainstorm with. So we chose myself, Director of Communications. We chose my colleague Nathan, who's our Communications Associate, who manages our social media. And we chose our Development Manager Valerie, who does all of our institutional grant writing. And these were the 3 people we felt could identify tasks that are repetitive, that require grunt work, where simple AI tools could help just alleviate some of the burden of repetitive tasks.

Tony Martignetti
Simple tasks. Very simple. Because we're after low-hanging fruit here.

Kate Dreyfuss
Yes, we're— and we're not trying to replace anybody's jobs. We're trying to just cut back on repetitive tasks that take a lot of time.

Tony Martignetti
Okay. So, Jen, jam session, step number 2 in the framework is, I believe, is that the jam sessions?

Jen Frazier
Jam sessions. That's right.

Tony Martignetti
Jam sessions. So after brainstorming, then jam sessions. What's the difference between a brainstorming session and a jam session?

Ben Freda
It's actually the same.

Jen Frazier
I mean, it's basically the same except for jam, you're eating, obviously. No, I'm just kidding.

Tony Martignetti
Did I mix up the framework?

Ben Freda
No, no, no. Step 1 is learn.

Tony Martignetti
Learn. Oh, learn.

Ben Freda
Yeah, yeah, yeah. See, just that.

Tony Martignetti
Well, you're all suffering with a lackluster host. And Kate was too polite to say, you're wrong. Yeah.

Jen Frazier
That's sweet.

Tony Martignetti
Step 1 is not brainstorm. Step 1 is learn. So step 1. All right. So Jen— That's okay.

Jen Frazier
They did the learning. She said they talked to Ben, they got some training. That was the biggest thing. Like, what is this and what does it do and what does it not do? Once you have that framework, then you can have your jam session because then you're like, oh, okay, these ideas are percolating. But to Ben's point, you're not immediately diving into Okay, how is all that stuff I just learned gonna help me? You actually stop and come back and go, what's going on in my job? What do I hate doing? What are the things that I take, like, just the sort of grunt work, the repetitive tasks? What's stuff that I'm like, oh my God, I hate sitting down with a blank slate looking at emails I need to go right back to all my grantees. I know they're gonna be, it's the same stuff every time, but it takes me a while to find everybody, do the things, write the note to everybody. Whatever. I have to go research. I have to go dig through papers. I do whatever that's like a lot of just sort of like, ugh, I hate that part of my job. So you don't even really get into like, can AI help me with this yet? Like Ben said, you're just really like, let's talk to everybody about their jobs. So the jam sessions are more about like, hey, let's just actually talk about your day-to-day and what's the stuff that sort of is a drag.

Tony Martignetti
Yeah.

Jen Frazier
That's your jam session. I know it doesn't sound like a jam session, but it is because it's, it's also very cathartic. Because you get to like talk it out with your, with your teammates and be like, oh, this is kind of a drag. And you're like, oh dang, okay, great. I can help with that. But let's not get to that yet. Let's just say, tell me what you actually literally don't love about your job. Just get it out. It's good therapy. It's blah, get it out. And then you can be like, great, I can probably help you with that.

Tony Martignetti
Okay. And it's New Music USA. So a jam session is appropriate.

Kate Dreyfuss
It could be EAM every day.

Tony Martignetti
All right. Yeah. Okay. So now we have our big list. We'll go back to Ben, back to Ben on the end. What's our big list step look like?

Ben Freda
So big list, there's the normal way of doing this, and then there's the technologically you get like an A+ way if you do it, which is to use AI to help you with it. Okay. We don't even want to go there necessarily.

Tony Martignetti
The normal way.

Ben Freda
The normal way is to listen to, to people in the jam sessions, and anytime an idea comes up, hey, I wish, I, I hate, like Jen, you know, like Jen said, I hate writing emails to all my grantees because they're all the same. There's a little bit different thing in each one, which is kind of annoying. You write that down. That's all. Just write it down. Make your list. Make a list. And so as you do that, by the time you've done these jam sessions, you've hopefully got a list of, you know, 15, 20 things on it. So that's the base. That's the normal way. Okay. If you want an A+, do the AI stuff.

Tony Martignetti
Yeah, do the AI version.

Ben Freda
Okay. The AI version would be to go to, you know, a Claude or something like that. Record all the meetings, by the way.

Jen Frazier
Yes, that's the key. Record them.

Ben Freda
So record the Jam Sessions. Then you download the transcript, right?

Jen Frazier
Yep.

Ben Freda
And you've got all the words in the transcript. And that's all these AI tools work on. They just work on words. So you've got all the words in the transcript. You take the transcript, dump it into Claude or ChatGPT, whichever one you want to use, doesn't matter which one. And you say, hey, I'm doing a jam session. And the purpose of this jam session is to find everyone's grunt work, right? The things they hate doing, the repetitive, the rote stuff that maybe AI might help with. And can you read through these transcripts and pick out ideas that you find? There you go. So you could do it the computer— and the A++ way is to do it both ways. And then combine them.

Jen Frazier
And compare. Yeah, compare and combine.

Tony Martignetti
Okay. Do we know yet whether all the items on our, on our, on our big list are AI achievable?

Ben Freda
We do not. No, we do not know. And a lot of times they won't be.

Jen Frazier
Exactly.

Ben Freda
So that's what that next— so then you go on to the next step, right? The prioritize step.

Tony Martignetti
Okay.

Ben Freda
So I don't want to skip ahead on my steps.

Tony Martignetti
Okay. Is there anything more you want to say about About step 3, the big list?

Ben Freda
No, I don't think so. I mean,

Tony Martignetti
we covered the big list. I did.

Ben Freda
The big list.

Tony Martignetti
Yeah. Kate, how many items do you remember? What was your—

Kate Dreyfuss
So, between— so, Ben met individually with me and my 2 colleagues, and between the 3 of us, we came up with a list of 15 possible use cases.

Tony Martignetti
Give us a couple of examples of—

Kate Dreyfuss
Sure. Well, the easiest one is an AI note-taker. But things like, so we publicize our grantees on our website. To do that, we have to create a lot of profiles for each grantee, so like 150 profiles per year in WordPress. So very simple repetitive task that takes a lot of time. That's one. An AI writing partner for social media strategy is another example. Let's see what else.

Tony Martignetti
One more. Sure.

Kate Dreyfuss
A knowledge base for grant metrics. So basically loading all of our existing grants into Claude and then making it easier for our grant writer to find metrics from past grants and update them just to save a lot of time.

Tony Martignetti
Okay. All right, cool. Now we're prioritizing. Jen, why don't you lead us through this step?

Jen Frazier
I think that, you know, honestly, that's a big one where you— I mean, again, the list that you're making usually isn't like, well, I'm going to do this in Claude or I'm going to do this in whatever. You just say like, I really hate it when She's already gone through all the steps, but when you're first doing it, you're usually just saying, God, I really, I need to, I need to go back through all of our past grant applications, or I wish I could go back through all my past, you know, applications and pull really great information out of that. God, I wish there was a way to do that. Then you figure out that AI could do that. So like we said, there's usually a list of things. You have to go through them and figure out which ones are actually going to be kind of the most AI-friendly, you know, like these are going to be achievable. But back to Ben's point, you're literally looking at like high impact, low effort. So you have your— think of a little grid, you're like impact over here and effort over here.

Tony Martignetti
One is X and one is Y.

Jen Frazier
One is Y, right? So you're looking at high end. So you then as a group, you kind of map them out and you go, how many people in the org will be impacted if we implement this solution? Yeah. That's gonna be a high-impact solution, and it's actually not gonna be that hard. Obviously, rolling out an AI note-taker is like the lowest effort thing you could possibly do, and it will have a huge impact on everybody in the entire organization if you roll it out properly. So that's the easy, like, boom boom. It's a double win, you know. If you go, I want to create some fairly complicated custom, you know, thing or whatever, you're like, ooh, that's a high effort, and it's only really gonna impact maybe maybe 1 or 2 people on the team. We're gonna put that in the let's get to that later list. Later on the priority list. So that's where you make a grid and you plot all 15 things out on that grid with maybe some help from then your consultants because you're like, I don't know how difficult this is to implement. And so that's where really the folks who've been doing this work can kind of jump in and be like, oh, that's actually a pretty high effort. It's not— I know all the Although marketing says it's really easy, but actually that's going to take a little bit more time than some of these others. So helping a group sort of sift those things onto the grid, it's easy to sort of pick out the top 2, 3, if you're ambitious, maybe 4 of those high impact, low effort.

Tony Martignetti
How long does this 5-step framework take to work through at an organization that has like New Music USA? You said 10 employees? 10. Okay. So how long did it take, Kate?

Kate Dreyfuss
Well, it's still very much in progress. I mean, you know, like everything, it's a matter of prioritization. So, I mean—

Tony Martignetti
When did you start then?

Kate Dreyfuss
We— when did we start? Was it— we really started in earnest in the fall, right? Yeah.

Tony Martignetti
So it's been 8 months or so? Yeah, roughly 6 months. 6 months.

Kate Dreyfuss
Yeah, I would say about 6.

Tony Martignetti
Yeah. Okay.

Kate Dreyfuss
Yeah, maybe even less.

Jen Frazier
And now you're at the point where you've got your list.

Kate Dreyfuss
Yes. So we are now starting the implementation phase. We have our, our top 4 that I alluded to, and we're gonna start with implementing 2 of those 4.

Tony Martignetti
Okay, so you're at the top 2 to 3, uh, and the blueprint— you're in the blueprint phase.

Kate Dreyfuss
Yes.

Tony Martignetti
What's, what's this blueprint all about, uh, Ben?

Ben Freda
Well, the blueprint is supposed to be more detail On each idea. And you want to capture what you're trying to replace. So what is the pain point that you're trying to replace? So what is the thing that people hate doing that AI is going to do? How is it going to do it for you? And hopefully a little bit about how it's going to be implemented. Again, you might need some contractor help or, you know, do some research on this, but a lot of times it's going to be, hey, we just have a custom prompt that we're going to put into a Claude. Mm-hmm. Other times it might be, hey, we need to connect Zapier with an AI system or something like that. But you just want a page or 2 on each idea. And the reason you want that is so that you can take it and hand it to somebody to implement for you. Yeah. So it's either maybe somebody on your team or in your organization, or it's somebody you need to hire from outside. But it's like at least a sort of a concrete, you know, not just what this— why this exists and what it's going to do, but a little bit about how it should be done.

Kate Dreyfuss
And if I may add, I think the Blueprint is also very helpful if you're in a position where you need to get leadership buy-in for implementing such a tool. Nonprofits are slow to adopt new technologies. Come on! And I mean, arts nonprofits, I think, are even slower in many cases. So, you know, having this Blueprint was extremely helpful. I was able to send it to our CEO and say, These are 4 tasks that we can undertake. This is how much time it's going to take to implement them and how much time it could save us in the long run. It's the ROI. Yeah, yeah, exactly. And metrics like this are so important, and the blueprint really helped establish those and helped me get leadership buy-in.

Tony Martignetti
And then it's implementation. Okay. You're gonna talk some about ethical issues around AI, I believe. That's what your session description is.

Jen Frazier
Responsible adoption. There's a little bit of responsible adoption.

Tony Martignetti
Yeah, okay, flesh that out for us.

Jen Frazier
I mean, essentially, there's— I think we kind of, because we only have a 30-minute session, we can't touch on everything, but we do talk about, at the beginning, a little bit of just about, again, like backing up to saying like, we understand there are a lot of considerations to go into using these technologies. There's a lot of focus right now on the environmental impacts of data centers and things like that. So, and the politics or the ideologies of the folks who run the various companies that do the large language models that we're using right now, the generative AI that most folks are using. So mostly we just wanna bring some of those ideas to the fore. We don't have a lot of time in our session to talk about it, but we can do a whole other— I mean, I'm sure there are other sessions. Yeah. This— at the NTC this year about the ethics around using the tools. So it's also— there's a lot of thoughts about, like, what does it still mean to be the original creator of something? What does it mean now when all the ideas of all of humanity are basically in a giant robot machine and it's— you can spit it back out at you? So there's all these different sort of ethical, theoretical, ideological considerations, environmental, all these things. And it's just important for us to help people understand. Yeah. In the learning part, we very much say, go to some learnings that actually really dig into some of these things. If you don't know about these things already, you should educate yourself. And then, like in a session I had on my Untangled a couple weeks ago, it was like, talk about these things as a group inside your organization.

Tony Martignetti
She drops the name of her podcast.

Jen Frazier
I did.

Tony Martignetti
But she did it so obliquely that—

Jen Frazier
It's fine.

Tony Martignetti
Probably not even noticed. I just called it out, of course.

Jen Frazier
Thanks. But, man— It's

Tony Martignetti
called Untangled.

Jen Frazier
Untangled. Brian Miller and I were talking about exactly these things to do before before you get going, which is to say, as an organization, where are we with our values and what kind of, you know, organizations do we want to work with, whether it's the AI tool or the vendor that, you know, prints our posters or whatever. We need to think about that as an organization. You probably already have those conversations about other vendors or other folks you want to partner with. Your AI tools are no different. So, it's like, how do we want to take our ethics and our values and make a values-aligned decision about what tools we're going to use and how we're going to use them as an organization.

Tony Martignetti
I like your analogy about working with other partners, consultants, partnering with other nonprofits. I mean, you would consider their values.

Jen Frazier
You would.

Tony Martignetti
You should do the same here. You should. That's a great analogy.

Ben Freda
Yeah. And let me add one thing about our framework.

Tony Martignetti
We're about to wrap up now. I want to thank Jen Frazier. No, go ahead. I got him. He's got him.

Ben Freda
You got him. He's like, wait, what? I was like, oh, my mic's off. I think I turned it off. It happens every time.

Tony Martignetti
Please, please add what you'd like to add.

Ben Freda
So framework-wise, by the way, you can actually download our framework from our session notes, from our session if you want. We have a little framework handout. Ethics is captured in part 5, okay, which is the blueprint and execution phase. As you do that, you want to, every time you implement a new tool, you want to update This thing, which you should have, called an AI use policy for your organization. Correct. And so the process of creating one of those should hopefully capture your ethical stance on this stuff. You should consider what you want to use, what you don't want to use, and put that into your AI use case or AI use policy. And every time you finish, you hit, you know, step 5 and you start executing on your ideas, you want to go back to the AI use policy and add whatever tool that you've created or adjust what you have in that policy to reflect the tool that you have adopted or the thing that you've created.

Tony Martignetti
Can we spend a couple minutes on the environmental impact? I mean, the environmental impact is unavoidable.

Jen Frazier
It is.

Tony Martignetti
As soon as you start entering prompts and then massive data centers are churning through your prompt and going through their quadrillions of bits of data and So the environmental impact is unavoidable from the get-go.

Kate Dreyfuss
Correct.

Tony Martignetti
If you are aligned with the sort of climate change mindset and that climate change is an issue for your nonprofit, even if it's not your work, but it's part of your values, Then, then how you have a, you have an unavoidable—

Jen Frazier
How do you reconcile that?

Tony Martignetti
Conundrum. You do. That's, you know, a dissonance.

Ben Freda
Yeah. So, so this is how I think about it. Everybody should think about it in their own way. Do their own research. Figure out, which is a terrible phrase to use, but do research. Figure out how you feel about it. So the way I think about it is there are environmental trade-offs about everything we do. So Drive a car, right? You should probably, if you can, take the bus, take public transit, not contribute to climate change, right? But there are good things about cars. There are ambulances, for instance, which can take people to a hospital. So it depends on what you're using it for, right? So do you want to use AI to write yourself jokes to make you laugh before bed? Probably not. Are you using AI to change the world in a better way? And that's going to be super impactful, then, you know, you want to think about it. The other thing to think about is the context of the environmental— the actual, you know, the actual data about what, how, how the, the data centers impact the environment is a bit opaque. That's, I think, the biggest problem with using these tools is that we actually don't really know how much energy they use and how much water they use. It's hard to extract that information from the, you know, the systems that are in place. On the other hand, we do know that Google's largest data center last year used as much water as the average golf course in the United States.

Jen Frazier
So it's the golf course analogy.

Ben Freda
So what's more useful? What's more, you know, what contributes more to the world? I would argue, you know, a data center, but maybe not. You know, it depends on your perspective.

Jen Frazier
If you're a golfer, maybe not. But I think about all the golf courses on the planet. And the places also where golf courses are, let's go with Palm Springs or Arizona, same thing. It's like ridiculous amounts of water that goes into a very luxurious, I think, elitist— For a few. For a few.

Tony Martignetti
For 1,000 people.

Jen Frazier
Yeah, elitist sport that's very expensive, blah, blah, blah. So, there's that. There's the environmental impacts generally right now are water consumption is the biggest thing. And obviously, but there's other impacts about like where are they building the Mm-hmm. Who are— who's actually staffing the data centers? What people are— what folks— what's happening in that community? A lot of folks are actually fighting the building of data centers in their local communities and things like that. So there's— you have to kind of like dig into it, but to Ben's point, it's very hard to find this information. They like to hide it from you as much as they can, but there are different ways. So I'll just drop another name. There's an organization here called Change Agent, which is an open source based and very, uh, an AI tool that's just like, uh, ChatGPT or Claude, but they're built from and within the social sector. So shout out to those guys. They will give you a rundown of how they are combating the environmental impacts of the data centers that they're using to process their stuff. And so they do a lot of carbon offsets. Exactly. They, they do a lot of work internally as an organization to already think about this and combat that. So they're, again, because they're from the space, they're already thinking about this and they're like, hey, we know this is a problem, we have ethical problems with it, but yet we also see huge benefits from organizations being able to use these tools. So they're a good one to also just look at if you're considering AI tools.

Tony Martignetti
Kate, I'm going to give you the closing word. How did New Music USA reconcile the the environmental impact with the return on investment that you— that it expects to enjoy from using the— yeah, picking up the low-hanging fruit with AI?

Kate Dreyfuss
Sure. I mean, I think the way we're thinking about it is more in terms of, you know, these tools are here and are going to be broadly used and adopted by most organizations and most people. So our take on it has been, how do we learn to use these tools that are here as responsibly as we can use them?

Tony Martignetti
All right. Perfect way to start— that's, uh, to end. That's Kate Dreyfuss, Director of Communications at New Music USA. With Kate is Jen Frazier, CEO and Founder at Firefly Partners, and Ben Freda, President of BFC, open paren, BFD, close paren, Digital.

Ben Freda
Whichever you prefer.

Tony Martignetti
No, the name of the company is BFC Digital. Ben Freda. All right, Jen, Kate, Ben, thank you very much.

Jen Frazier
Thank you, Tony. Thanks, Tony.

Tony Martignetti
And thank you for being with Tony Martignetti Nonprofit Radio coverage of the 2026 Nonprofit Technology Conference.

Kate Martignetti
It's time for Tony's Take 2.

Tony Martignetti
Thank you, Kate. We have a new character in our ensemble at the gym, community gym that I go to. This is Sam. Now, Sam is the gentleman who invited me to sign Jerry's birthday card. Remember Jerry, who I don't know, but I recognized him when Sam told me where he sits because he sits at the same bike every single time and he's turned 91. I signed his card. That was Sam who invited me to do that, not because Sam knows my name, but just because he sees me a lot. So he invited me to sign. Sam is, uh, Sam is interesting. He's very, uh, talkative. Um, he's, he's, I think he's north of 80. He's got a, this great shock of white hair, all pure white hair. Wow. Slender, tall. I believe he was in the military. I think I've heard him say through the years that I've been going, I, I, I, he was in the military. He may have even flown, like a Navy, maybe a Navy pilot or something, or Navy flyer. Not necessarily pilot, could be backseat some aircraft. But, um, I have to, I have to get a little more bearing on, on Sam. So I just, but I just learned his name. Yeah. That's why I'm introducing you to him. I didn't know who he was, but I heard him introduce himself to someone else. Yeah, he's chatty. Like, he walks around to the different stations and different machines and people, and I don't know if he's interrupting, but he's imposing. You know, he wants to talk about— he talks about sports, he talks about politics. He talks about town stuff, like little events that are coming up. You know, like we had the— we had St. Patrick's Day last month festival, things like that. But he goes around and he is slender, but it's not because he works out a lot, because he does more walking in the fitness center than he does actually working out at machines or weights or anything. He's, um, he, uh, he likes to, he likes to tell stories and talk to people. So that's Chatty Sam. Very nice guy. Well, if you can put aside the photos, aside the fact that he's imposing on everyone's workouts. He doesn't impose on mine, uh, and because I'm on the elliptical or I'm down on the floor and I'm panting as I'm doing my planks. And so— Plus, he'd have to bend down to talk to me. So I'm inconvenient. But he talks to a lot of people unless you're— I noticed unless you're on the treadmill, the treadmill he avoids. But if you're at any of the machines or even on the bikes, he comes over for a chat. It's a little much. A little much, I believe. That's Chatty Sam. And that's Tony's Take 2.

Kate Martignetti
Kate. So do we know if the guy ever got his birthday card?

Tony Martignetti
Oh, Jerry did get his birthday card. Yes. But it was about 2 weeks late because he didn't come in at the same time Sam was coming in with the card. So by the time they got together, it was about 2 weeks after Jerry's birthday.

Kate Martignetti
It's still the thought that counts.

Tony Martignetti
Absolutely. Absolutely. And he's probably still— I'm sure Jerry is still scratching his head. It keeps him up at night. Who's the person who signed Jim Rat? Who could that be?

Kate Martignetti
Who's Jim Rat? We've got just about a buttload more time. Here is Ethically Using AI.

Tony Martignetti
Welcome back to Tony Martignetti Nonprofit Radio coverage of 26 NTC. That's the 2026 Nonprofit Technology Conference. We are all gathered, all these technologists and technology users in Detroit, Michigan. With me now are Chris Rosica, CEO and President at Rosica Communications, and Bethany Friedlander, President and CEO at New Bridge Cleveland. Chris, Bethany, welcome.

Chris Rosica
Thanks, Tony.

Kate Dreyfuss
Thank you.

Tony Martignetti
Pleasure to have you Welcome to both. Your session topic is 5 Tips for Ethically Using AI to Develop Compelling Website and Social Content. Chris, I'm going to ask you to just give us a high-level overview to start us off.

Chris Rosica
Sure. Our room was pretty packed with mainly nonprofit executives who were really, based on the title of our topic looking for ideas and how they could use AI to generate content, whether it's for their— for an article or a blog, how to repurpose that content for social media. Also how to leverage earned media coverage, say an interview with Bethany Friedlander from New Bridge.

Tony Martignetti
And take that interview and turn it into, say, an e-newsletter for stakeholders.

Chris Rosica
So it was really about utilizing AI in ethical ways, of course, but utilizing it to really integrate and diversify marketing communications so it's less taxing on staff to do that work. And it makes it more feasible to do the work that in the past may not have been a priority based on resources, resource limitations, and lack of qualified personnel to do that work.

Tony Martignetti
All right, thank you. That's ideal. Thank you for that. Now, Bethany, are you a client of Rosica Communications? Is that the relationship here, client and consultant?

Chris Rosica
Yes, we actually met at South by Southwest EDU. We sat next to each other at breakfast and have developed a really strong relationship, primarily working on elevating the story of New Bridge Cleveland. We're a small nonprofit, and so looking to frame that story, how do you become not a nice-to-fund but a must-fund, and how do you do that particularly bringing national money into Cleveland, which is not a place where a lot of national money finds its way, actually.

Kate Dreyfuss
We have a very robust local philanthropy, but not nationally.

Tony Martignetti
Okay, and before working with Chris and Rosica, what was your experience, if any, with artificial intelligence using?

Kate Dreyfuss
I think we were probably in the same place a lot of nonprofits are, which was, is this cheating, right? We've written— we answered the same grants 15 years in a row with the same questions They've never changed the application. Is it cheating to use AI? And I would say no. We have our own blinders on about the language we use to discuss ourselves, and I think it's important to have an opportunity to have that looked at, challenged, and improved.

Tony Martignetti
Okay, Chris, why don't you walk us through— you have 5 strategies of ethically using AI for nonprofits. Web and social content. Can you talk— could you just tick those off and then we'll talk through?

Chris Rosica
Sure, sure. So the ways that you can use AI include positioning and messaging and help frame— to frame the story. The second is for repurposing content for stakeholder

Tony Martignetti
communications. Communications. The third way that you can utilize AI is to help you expand your website content, and that includes articles, thought leadership articles on your site. The fourth is to utilize these tools to help you Be more searchable online. So AI search today, Tony, is huge, and many nonprofits were struggling, are struggling, or still struggling with search engine optimization. Now you have AI search, and how do you rank? How do you rank in AI search? So that was another key topic we spoke on. And then the last is really creating your brand persona. And How do you stand apart? How do you differentiate your organization? Okay. Okay. Around the ideation, you know, thinking about brainstorming how you might have started using AI at New Bridge Cleveland, Bethany, what did you— was there a brainstorming session or meetings or like how did you decide where Where the best use cases would be to begin?

Bethany Friedlander
So a number. So first of all, a lot of different opportunities, right? So one of the biggest for me was character limitations in grants. So you can write the most eloquent answer in the world, and if it doesn't fit within 250 characters— so it's fantastic as a tool to take you and edit as an editor, also to help you brainstorm. So the opposite, right? So sometimes I know exactly what I want to say and I'm not saying it in a few enough characters, and sometimes I don't know what I want to say at all. And so I can start in a very casual way throwing ideas and words into AI and then getting back the brainstorming that then I can then further refine.

Tony Martignetti
Chris, any thoughts on I think

Chris Rosica
that having AI really get

Tony Martignetti
to know you and what you're all about is critical. And Bethany's done a great job of doing that, of educating the AI, of prompting it, of having it really explore what her brand is all about. Most websites today aren't up to date with people's messaging. Why? Because all All organizations evolve. All nonprofits are evolving. The landscape is very— it's challenging right now because federal funding has been cut across the board. And how do you stand apart and diversify your funding mix, right? And that's one of the things that NewBridge has done a great job with. It's really giving AI all of the information And they've also updated their website too recently with very current messaging. Again, many fall behind in that area, not just nonprofits, but across the board. Yeah. Corporations do as well. So that's something that I've seen really pay dividends to New Bridge. And Bethany, I'm sure, has a lot to say about how she works to inform the AI and really let it Yeah,

Bethany Friedlander
so I throw every email that I write into it with the quote, with the prompt of make this better. I may or may not use the change, but every time I do that, it's learning my language, it's learning my phraseology, so that then when I ask it to create something new, it's creating it more and more frequently and tighter and tighter, closer to my own voice. We're also a fairly complicated nonprofit in terms of how we do the work. And that's been a challenge since our inception to explain to funders. And so allowing myself to engage in a dialogue with AI to say which parts of this are the most resonant with the audience that I'm trying to get to, and understanding that it really only matters that it's heard. It doesn't matter what I'm necessarily prioritizing. So letting it do some of that work for me. and acknowledging that I may not be the best lens, we're very close to our work. I don't think there's a nonprofit here who isn't deeply passionate about their mission, deeply resonant. That can be a huge limitation to your success because it puts huge blinders on you and the language you use to describe your work.

Tony Martignetti
That's interesting, yeah. When you're uploading everything, including your emails, are you doing this to a learning basis? that is unique to New Bridge? It's not— this is not in the public domain,

Bethany Friedlander
right? Well, and that's a funny question because I am giving it a tremendous amount of power and information that it can use then to improve other nonprofits, right? But I think it's still worth it to have that refinement too. And I would say now, I would say about 85% of what it writes, I clearly hear my own voice. I no longer question it. It does not seem like a foreign entity, but that's also because I have my own account. I don't let anyone else use my account, so it's only getting input from me. It's not getting input from the entire organization.

Tony Martignetti
Okay, okay. But it is one of the public models.

Bethany Friedlander
Yes, it is. It is.

Tony Martignetti
Okay. And so that doesn't— I mean, every email, maybe not every email literally, but, you know, if it's 95% of your emails or even 75% of your emails. That doesn't bother you that they're now in the— is it right to say they're in the public domain? Like, they're in the domain that this large language model, which— what do you use?

Kate Dreyfuss
ChatGPT.

Tony Martignetti
ChatGPT, okay. That ChatGPT is learning from. Yeah. That doesn't trouble you?

Kate Dreyfuss
It does.

Tony Martignetti
Suppose somebody does a prompt, write me an email in Debbie Friedlander's tone to have her director of finance write a check for $50,000 to my company?

Bethany Friedlander
Yeah, no, I think it's a fair question. And we had a great audience member who wanted to talk about, like, the ethics of all of this, and I think it's incredibly important. I think where my head's at right now is a 15-person staff with a $2.5 million budget that has to be raised like clockwork starting July 1st of every year. And right now, the benefits outweigh the liabilities. I care deeply about my participants and making sure that no private information about my participants is ever included in those emails. I would never do that. I would also never put an email in that had any private information about a funder or an arrangement with a funder. Yeah. But if we're talking about, you know, conversational emails where I'm talking about New Bridge, and every day I have 10 opportunities to describe the organization, they don't all land equally. And so being able to run it through that and recognizing, I think it's elevated the language. And my proof is, is that it's elevated our status in Cleveland. It's increased the number of tables that we're sitting at. It's increasing the number of dialogues we're participating in. And I think ultimately that pays off.

Tony Martignetti
That's interesting. So you're seeing those kinds of outcomes? Absolutely.

Bethany Friedlander
Absolutely. Being a thought leader. And I mean, you can only be a thought leader if your thoughts are well organized. And I don't know that that's something that naturally we all do particularly well.

Tony Martignetti
And heard.

Bethany Friedlander
And heard.

Tony Martignetti
Also. Otherwise, if your thoughts are well organized and nobody reads them, that's called a diary. That's not what you're aspiring to. Right. Okay. No, interesting. But you're right.

Kate Dreyfuss
There's risk and there's And you are enriching something that you're never going to get back, so you have to hope that your returns come in another way. And I think right now benefits outweigh liabilities. I can't say that that's going to be true forever.

Tony Martignetti
Okay. Chris, say something about the AEO, artificial engine optimization. You said earlier a lot of nonprofits are still working on SEO, but that really has been Overtaken. It's almost an anachronism now because the artificial engine, the AI engines are generating summaries that don't lead to your website. How do we overcome this?

Chris Rosica
Well,

Tony Martignetti
certainly

Chris Rosica
there are a lot of pretty simple and straightforward ways of understanding how AI

Tony Martignetti
works. And it's not really reverse engineering either. You can look at who is ranking for certain key phrases. So if you put in workforce development, healthcare workforce development organization in Cleveland—

Kate Dreyfuss
We better come up first.

Tony Martignetti
You better come up first. Exactly.

Kate Dreyfuss
This relationship is not working otherwise. Exactly.

Tony Martignetti
You've been admonished.

Chris Rosica
Yeah. So if you put in a certain key phrase, depending on where you're located, and the service, the geographic service area that you're serving, right? You can be a national organization. Most of the time though, national organizations have a local presence, right? So the Very— the telltale signs are,

Tony Martignetti
you know, are you ranking in AI search results? And if not, look, scroll down, and then you'll see people who are searching AI also ask, and then you'll see some additional key points. So you'll know the important— that you'll know the ways people are searching in AI for answers. After you understand that, then it's time to modify your website and the content

Chris Rosica
on your website to create summaries, really succinct summaries, and Q&As or FAQs. FAQs, yeah. That address those questions that are being asked on a regular basis. So that's— these are just some very basic ways of beginning to appear in AI search results. There are some other things that are a little bit more technical. So if any of your listeners have questions about that, we're very transparent and very willing to share with them those specifics. So if they have a web developer that's part-time or an agent, a company that's working for them, we can steer them in the right direction. Okay. in the right direction and let them know what they need to do so they rank in AI search results.

Tony Martignetti
Okay. Bethany, how about expanded web content? I guess this is generating website content for you as well as social, right? As well as social content.

Bethany Friedlander
Yeah. So before we met, Chris, we had no social media presence. I just didn't have anybody. I feel— well, let me take a step back. I think it's really important if you're going to add anything to your organization, you need to be doing it consistently. So having, you know, a Twitter account, if you're tweeting or Xing— Xing, I guess? Xing, I suppose. Xing once a month is irrelevant. So I never had the staff to do it. Chris has showed us how we can take content that we put in one place and use it in multiple other places in a really simple way. And so we've greatly expanded our footprint. And I think It's really about— it's not about eliminating work, it's about doing work smarter. And so I think we were in this phase of just, we can't, and now it's, we can, and we can do that smarter.

Tony Martignetti
Is that AI-related? It is.

Kate Dreyfuss
So explain how you took— you can take like one of my articles and then—

Chris Rosica
So Bethany was featured on Medium, Authority Magazine's Medium platform. Millions of people have exposure to that, but how do you know if the right people are, you know, tuning into the website and seeing Bethany's bylined article or op-ed piece? So what we did is took—

Tony Martignetti
utilizing AI, took that interview with Authority Magazine, turned it into a blog post that went onto New Bridge's website, turned it into a LinkedIn social post, All in an automated fashion that saved so much time. And there's no— it's in the information. The article was in the public domain to begin with. It's on medium.com. So I understand the question you asked earlier, Tony, about sensitivity and do you really want to give it confidential information. And 99%— I mean, almost all of what we do is Not that at all. We're very wary of doing that, and I concur. I get the sense that you feel you have some concerns about that, and we have the same concerns. I know New Bridge does too. But in this area, that article now is turned into all this content for different channels, and it's all unique content. Same holds true not only for the earned media coverage, but for an article, a thought leadership article that is written for the New Bridge website that can be turned into a media pitch. That can be turned into

Chris Rosica
a, into social posts for all your social channels. And then we use AI on an ongoing basis to take one social post that We write and turn it into social posts for all of a nonprofit's channels. So it's unique content, but it's still conveying the same message.

Tony Martignetti
You've had your session already. Yes. So what are some of the questions you got that struck you, that are memorable?

Bethany Friedlander
So we, we asked them a question, which was to put a simple question in into whatever AI they're using, which is, what movie star is my organization? And I do that, and I talk about that because it allows you to see how you're being seen. It's the language you're using. So interestingly enough, overwhelmingly in the room, everybody got Tom Hanks. And the

Kate Dreyfuss
inclination was, oh, then AI is not that smart. It only has one actor. And my response is, no, that's because we're all talking about our organizations in the same way. Right. talking about ourselves as kind, good people. Now, interestingly enough, New Bridge doesn't come up that way. New Bridge comes up as Viola Davis, and that's partly because we talk a lot about trauma, resiliency, and grit. And if you want to talk about somebody who embodies that, Viola Davis really embodies that. So my suggestion to people is, if you don't like what you're getting back, then think about how you're talking about yourself, because we're really much more complex than just being good.

Chris Rosica
By the way, Rosica Communications is George Clooney. So, which, which was a shocker, by the way.

Tony Martignetti
Yeah, Clooney. Okay, now I'm thinking of Alec Baldwin. Yeah, I was thinking of the personal side of Alec Baldwin. George Clooney. Yeah. So are you happy with that?

Chris Rosica
Very happy. Yeah.

Tony Martignetti
Okay.

Chris Rosica
Yeah, great under pressure. He always seems to be great under pressure. His persona, right? His acting persona, very consistent. You can rely on him. Yeah, those types of attributes.

Tony Martignetti
Yeah, does similar kinds of roles too.

Chris Rosica
Yes.

Tony Martignetti
Right. Okay. Leading man. Very good.

Bethany Friedlander
Okay.

Tony Martignetti
What's another question you got that might be interesting for folks?

Bethany Friedlander
So we also ask, what's something about my organization that's confusing or might be misunderstood by funders? Again, something that you may not see. It might be a huge blind spot for you, but can help you with with clarity, and then also what is the number one thing we could be renowned for? What can we be world best at? So I think asking it those questions, it's not, I mean, it's not foolproof, but it does give you a sense of how somebody who's not with you every day might respond to you, think about you, and what they might be left with as an impression. If it's not the impression you want them to be left with, change the language you're using.

Tony Martignetti
Yeah, Tony, we always say perception is reality. And how are you effectively demonstrating the value you bring to the stakeholders that you serve?

Chris Rosica
And if you're not differentiating your organization, and if you're not saying things that truly resonate with those stakeholder groups, then you're missing a huge opportunity to improve communications in order to serve more people that you're looking to serve.

Tony Martignetti
How about we leave it there? Thanks. That's Chris Rosica, CEO and President at Rosica Communications, and Bethany Friedlander, President and CEO at New Bridge Cleveland. Chris, Bethany, thank you very much. Thank you. Thank you for being with us, and thank you for being with Tony Martignetti Nonprofit Radio coverage of the 2026 Nonprofit Technology Conference.

Kate Martignetti
Next week, we've got more on AI from 26 NTC with AI for the Rest of Us and Your AI Acceptable Use Policy. If you missed any part of this week's show,

Tony Martignetti
I beseech you, find it at tonymartignetti.com.

Kate Martignetti
Our creative producer is Claire Meyerhoff. I'm your associate producer, Kate Martignetti. The show's social media is by Susan Chavez. Marc Silverman is our web guy, and this music is by Scott Stein.

Tony Martignetti
Thank you for that affirmation, Scotty. Be with us next week for Nonprofit Radio. Nonprofit Radio, big nonprofit ideas for the other 95%. Go out and be great. Okay, a little note here. We've got more. The way you said more made it sound like it's too much. It lost the effect. We've got more. All right, I'm exaggerating. You didn't do it that way. That much, but just— we've got more on AI. Please just do next week for me.

Kate Martignetti
Next week, we've got more on AI from 26 NTC with AI for the Rest of Us and Your AI Acceptable Use Policy.