Community IT Innovators Nonprofit Technology Topics

Nonprofit AI Management Advice with Mimi Yeh

• Community IT Innovators • Season 7 • Episode 39

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 26:16

Carolyn Woodard explores the gap between AI ambition and organizational readiness with Mimi Yeh, Engagement Director at PTKO, a strategy and change management consulting firm that works exclusively with nonprofits and mission-driven organizations. Mimi focuses on the people side of technology adoption — and when it comes to AI, she sees a consistent pattern: leadership is excited, staff are uncertain, and the gap between those two realities is where AI initiatives stall.

The conversation reframes AI adoption not as a tools problem but as an alignment and change management challenge. Mimi and Carolyn discuss how nonprofits can build the internal structure — governance, policy, shared understanding — that makes AI adoption sustainable rather than chaotic, and what funders can do to support that work more effectively.

Mimi and Carolyn discuss:

  • Why the biggest obstacle to AI adoption in nonprofits is not the technology itself, but the gap between leadership ambition and staff readiness — and how governance can close it.
  • How to use your organization's existing skepticism about AI (around climate impact, job displacement, or data risk) as a productive starting point for building policy rather than a reason to stall.
  • Why acceptable use guidelines for AI should not look very different from governance frameworks nonprofits already use for other tools and technologies.
  • How funders can shift from pressuring nonprofits to adopt AI quickly to supporting the capacity building and literacy that makes adoption meaningful and lasting.
  • What it actually takes to build an AI tool for internal nonprofit use — and why the human input, voice, and values behind that tool matter as much as the data fed into it.

Resources Mentioned:

_______________________________
Start a conversation :)

Thanks for listening. 


Mimi Yeh

If I can just give some like overview thoughts around AI and how nonprofits you might think of using AI and where there are some traps that folks can fall into, and organizations can sometimes have these pitfalls and places that get them stuck in a space and how we can kind of help them avoid that.

Carolyn Woodard

Welcome everyone to the Community IT Innovators Technology Topics podcast. I'm Carolyn Woodard, your host, and I'm very excited today to be talking with Mimi Yeah from PTKO. So, Mimi, would you like to introduce yourself?

Mimi Yeh

Yes. Thank you, Carolyn. And thank you for having me on the podcast. I am an engagement director at PTKO. Uh, we are a small consulting firm. We have presence in the Metro DC area, as well as in New York, and uh one staff member out all by themselves on the West Coast, uh, as well as some individuals in North Carolina. Um,

Mimi Yeh

We are a strategy consulting firm, and we do also help out with some technical aspects of uh work that is important to nonprofits and mission-driven organizations. And that can include things like CRM selection or other software selection, building out roadmaps that help to improve the tech stack that organizations use as an enabler of achieving their mission and their vision and their goals.

Mimi Yeh

I specifically focus on more strategy, human capital, and change management efforts. And the way that I usually describe it to people is process and technology are incredibly important. They are huge elements of what make an organization work well. But until we are all taken over by machines and robots, we have to have the human aspect uh considered just as much as the other two. So I really focus on the people aspect of changes and adoption of new capabilities and new technologies.

Carolyn Woodard

And I think in the nonprofit sphere and in philanthropy, but in nonprofits, that it's all about people. Like we're people organizations and we care about our communities and the people in our communities, and we have all these relationships and long-standing relationships. And um, yeah, so I'm really interested to hear um kind of your experience.

Carolyn Woodard

Can you give us a little bit of an overview on AI and what you're seeing?

Mimi Yeh

Yeah, and I'm really glad that we're having this conversation because so far, you know, what we're seeing is that AI can go in a hundred different directions really quickly. So what we've been observing in the past, I would say eight months, but maybe even longer than that, with the inception of AI really making a more specific and significant presence in nonprofits and organizations, is that organizations are not lacking an awareness of AI at this point. You know, if anything, there's too much awareness that's going on. Um,

Mimi Yeh

I think that a lot of leaders are hearing things like AI is transformative and you need a strategy and you need to act now. And then staff for their part are sitting there and thinking, well, wait, what am I actually supposed to do with this? And am I even allowed to use it?

Mimi Yeh

So I think, Carolyn, what's happening is that there's this growing gap between leadership ambition and staff readiness. And when that happens, that's usually when new and exciting initiatives can tend to stall. Um, because from the way that we look at things, AI really isn't a tools problem or a tools challenge. It is an adoption and an alignment challenge. So it's really about having a shared understanding of both the problem statement, but also the opportunity. Um,

Mimi Yeh

I think it's about having some clear guardrails for the organization, about what AI means to them and how they see AI as an enabler, not the final end-all be-all, but as an enabler. Um, and then it's really about having confidence across the organization about what to use AI for, how to use it, and that it's okay to experiment and try out some new things. Um,

Mimi Yeh

I think the teams that are making progress are not necessarily the ones that waited and picked out the quote unquote best tools. And rather, they're the ones that created a safe and structured way for people to engage with AI in a space where it makes sense for AI to play a role.

Carolyn Woodard

It's definitely what we're hearing from a lot of our clients as well. And I think it was very um this gap, like it's very deceptive because we're finding that AI requires a lot of change management. Yes. But it was sold to us all as you just ask it a question. Like there's no upfront anything. You it's already you have it in your tool and you just ask it and it tells you. And it's so it's like kind of wormed its way in as like an easy solution, but the amount of change management around it to be really intentional about it is just kind of invisible. So I love that you're you're bringing that up.

Mimi Yeh

Yeah, yeah, thank you. Yeah, and I think that's right. I think that with many new products, they have this wrapper around them that are described to nonprofit leaders and different individuals as it's the easy button that you push and it's gonna fix all your problems, and you don't have to worry about anything.

Mimi Yeh

And we've seen this with other technologies in the past, and we've also seen how difficult and complex it is to manage and operate in a space that doesn't have some basic governance, right? Some standards, maybe even some ethics and policies that are in place. It's not intended to over police the use of AI within your organization, but it is really intended, again, to give staff members and leaders a common understanding of what AI means to get your work done and how to engage with AI in a way that's going to be consistent with the organization's values.

Mimi Yeh

I will say one of the consistent messages or uh pushback areas around AI that I hear from a lot of nonprofits is oh, it's so detrimental to the environment or you know, it has such a climate impact in a negative way. How can I make it okay, right? How can I feel okay using a tool that has such a you know damaging um effect on the climate and on the world that we live in? Um, and

Mimi Yeh

Again, it comes back with that bigger question of let's not treat AI as a separate standalone item that has to face a different set of scrutiny and criteria. It should be the same consistent criteria and evaluation factors that are used when you're looking at everything in your tech stack, or even everything in the way that we operate and live and breathe. And do we take planes instead of cars? Do we eat beef instead of chicken? And you know, there's no um in

Mimi Yeh

There's no intention here to say that one item is good and one item is bad. It's more to say, let's take a macro view at all of those different choices that we make. Uh, and where's the right balance where maybe there's an impact on climate, and there is also an end result that significantly improves efficiency, operations, effectiveness, or really ability for that mission-driven organization to achieve the outcomes and the results that they hope to.

Carolyn Woodard

Yeah, someone said it to me about yes, we are hearing a lot about the energy use, for example. Yes. Um, but she said, like, as an organization, have you been concerned about how much electricity you use in the sense that, like, oh, we're just not gonna use electricity today? You know, like it's one of those things that uh it's all around us and it's in everything, and you can't operate as a modern functioning nonprofit without having lights on.

Mimi Yeh

Yes.

Carolyn Woodard

So I think that it's that was a good analogy because I think AI is gonna be in a similar place of it's just gonna be, you know, or are you gonna go to the well to get water to wash your clothes? Like we all have pipes and plumbing, so it's kind of similar to that. I think it's gonna be in that role. It's getting, I mean, it's not there yet, but it's getting there.

Mimi Yeh

Yeah, yes, absolutely.

Carolyn Woodard

So are there other reasons that nonprofits are skeptical about AI aside from the environmental one, which is really landing with people? I mean, there are lots of things to be skeptical about, but that one seems to be taking a big part of people's mental space.

Mimi Yeh

Yes, I do hear a lot of, I think there's this kind of blend of skepticism and concern. Um, the sustainability and the energy and environmental impact is definitely pretty close to the top of the list.

Mimi Yeh

And then the other concern that comes across is this idea of job replacement, right? This idea that it will uh eventually uh create a job loss and that will have such a negative impact on communities, on the nonprofit industry as a whole, or the philanthropic field as a whole. And um, you know,

Mimi Yeh

I think that it's skepticism sounds negative, but I think it can be looked at in a slightly positive way, only in that it helps to highlight the kinds of questions that organizations should be thinking about and should be able to address, you know, or or take some time to have a plan for.

Mimi Yeh

So when people ask, you know, about what's going on, when they ask about equity or the long-term consequences, um, they're really thinking about, you know, what happens to our data, right? It's is using AI in alignment with our mission? Um, are there ways in which the use of AI might create some harm? Um,

Mimi Yeh

Those are all legitimate questions. Those are also great feeders and starting points to the whole structure and the governance around AI that we were just talking about earlier. Those are great ways to shape out the policy that organizations would like to have, um, the practices and guidelines that they'd like to put in place, and maybe some security measures about what type of data can is and is allowed to be fed into an AI tool or an AI bot. What kinds of things might this organization decide we will never put information in there?

Mimi Yeh

And also when it we think about the alignment to the mission, it can get into things like what's acceptable use of AI versus what might we as an organization decide is not acceptable for us. Um, I think in the event that you and I went to, we listened to a marketing and communications lead for a nonprofit organization. And he said that one of the policies that his place has is they will never use AI to generate images that tie to their mission that would get displayed and made public on their website. And I think that's a great idea and a great example of identifying the guardrails and the guidelines around use of AI, right? It can do all kinds of things, right? You have nano banana, you've got all kinds of technical tools that have gotten much, much better at generating very lifelike images.

Mimi Yeh

And that director understood that to create an image, to generate an artificial image of a mission-driven event or population or constituency is something that carries so much risk and can have such a bad impact on that organization's brand and their value and their entity, that that organization has made very deliberate steps to say that is not okay. Um

Mimi Yeh

I think it's by addressing a lot of those big questions that hopefully that skepticism starts to lower and people get more excited about the productive and really valuable use of AI.

Carolyn Woodard

If you have leaders who are still on that skepticism side and maybe more reluctant, I agree with you. Like caution can be a good thing and being fully informed, you know, is a very good value to have. So like skepticism can be kind of coupled with those values.

Carolyn Woodard

But if you have someone that is really reluctant, um, do you have practical ways to get started? Like what kind of advice do you give leaders who just don't even know where to start with a governance or policy?

Mimi Yeh

This is gonna sound trite, um, but I do think that it's really the right answer. Thinking about governance for AI ought not to be very different from thinking about governance for other technical tools and other solution sets within your organization, right? The reason that the organization exists is to achieve a particular mission or to support a particular constituency. Um,

Mimi Yeh

Thinking about it from that standpoint and understanding the ways in which different tools should be enablers of getting things done and contributors. It's not AI is never going to be the silver bullet, or it should not be the silver bullet, nor should any other major technology out there, right? It is the direct services, the processes, the advocacy, the ability to convene a strong network and community of individuals who can all contribute to an outcome. That is the nature, I think, of many of the nonprofits' work that's out there.

Mimi Yeh

So creating guidelines around how to get that work done, that should be the central focus and element of an organization. And then it's really pressure testing the different possible uses of AI against those existing guidelines and guardrails. And I think that the solutions and ideas that organizations already came up with when it comes to big data, when it comes to LLMs in general, and when it comes to other even using a CRM, you know, there's particular policies around acceptable uses of different technologies and tools. AI really should not be that much different to it. Um I think that, you know,

Mimi Yeh

Once an organization has a good sense of how they want to use AI for what purposes, and then thinking through, you know, what are the risks that might be associated with each one of those? That's I think a good starting point to building out the guidelines and the SOPs and the policies of use.

Carolyn Woodard

Yeah. So whatever process you would use for other types of governance, that you talk to your board or you have a committee or you make a draft and share it, or however you get started, just get started.

Mimi Yeh

Yeah. And I think that, you know, when we link back to our earlier conversation around skepticism in the use of AI, I think a lot of times that skepticism comes from a place of risk management, right? And it comes from a place of concern that AI, my God, it's so powerful, it could do all these things.

Mimi Yeh

I think a really useful practice is to list out all of the risks and concerns that people have today, rather than shutting them down and rather than saying, well, that could never happen. I mean, it could, it could happen, but let's list all those risks out first. And then again, similar to many other organizations, right? You would then um evaluate what is the impact of each of those risks if they came to realization.

Mimi Yeh

If it transitioned from a risk to an issue, how big is that impact going to be? Whether it's on the organization, on the constituencies, on partners, on whoever it is. What's the impact and what's the probability? What's the likelihood that something like that could happen? And then what I would do is probably take the top risk items that have high risk or high probability, high impact, or medium probability, medium impact, you know, probably those top sets, and then start thinking through what type of governance structure, what type of policies, what type of guardrails do we want to put in place to help mitigate or reduce the likelihood that that risk could happen.

Carolyn Woodard

I have another question for you, kind of related to this, which is what are you seeing from funders or what are you hearing from your clients that they're hearing from their funders?

Carolyn Woodard

Because I'm getting this sense, we you talked earlier about the silver bullet and kind of some of these unrealistic expectations of AI as a transformative agent. And I am hearing that from some clients and some funders, you know, there's these grant proposals out now, request for proposal of how are you going to use AI, you know, to do your mission completely differently.

Mimi Yeh

Yeah.

Carolyn Woodard

And I'm wondering, are you hearing that too? And how are you advising um clients to interact with their funders around this?

Mimi Yeh

Yeah, we are hearing that a fair amount. And I I understand and I can empathize with the point of view of the funders, right? I think that they're looking for ways to, I don't know, exponentially or significantly help nonprofits achieve their mission. And

Mimi Yeh

I do think that they can consider AI as, you know, it's like a steroid shot that just suddenly rockets that nonprofit into the high-performing space or just gives them something that turbocharges everything that they do. Um,

Mimi Yeh

I think that there's a danger of unintentionally creating pressure on that nonprofit. Um, it's like they're being told you should be using AI, you should be innovating, why aren't you doing this? Everybody else is doing this. And

Mimi Yeh

I have to say it, you know, without sounding like a broken record, it comes back to if you don't have the right support structure in place, it can lead to some rushed decisions being made, it can lead to unclear policies and therefore people unintentionally going rogue or do taking on practices that might be contradictory to what leadership had hoped it to be, or honestly, people can just feel overwhelmed by it all. Um,

Mimi Yeh

I think that a more helpful role that funders could play is providing the ability and the resources to help organizations define their basic governance. I do think, you know, without sounding super boring, because I do think that there is a space to actually use AI, and we are seeing it with one of our clients um supporting experimentation, not perfection, not you know, getting the end-all be-all right in stage one, but trying out use of AI and understanding its potential in some structured use cases.

Mimi Yeh

And by doing that level of experimenting and piloting, working alongside that tool development to also think about what is our position on AI, right? What is our point of view around AI specifically for our organization? Organization. And then I do think that it's just as much about funding capacity building as it is about funding the tool.

Mimi Yeh

So hopefully, funders will understand and will support the nonprofit's ability to put together the adoption structure to maybe provide some investment in training, but not in training on how to use AI as a tool, but how to understand AI as an embedded part of the organization and the way that it can operate. So AI adoption, it's a capability. It's not a one-time project. It's going to keep happening, it's going to keep evolving, and it's going to keep needing support and resources to make it happen.

Carolyn Woodard

Yeah, I heard it said that I'm hoping also that there'll be more funding toward AI literacy because, as you said, like making good decisions, you have to know what you're making a decision about.

Mimi Yeh

Right. Yep. That's that's right. Um, that's right. And, you know, one of the the client that I was rec referring to earlier, we are helping them to build an AI tool that can be used internally by the staff in their organization.

Mimi Yeh

There's more to creating an AI tool than giving it a corpus of hundreds and hundreds of artifacts and documents, right? There's a voice to that AI tool. There is a personality to that AI tool. And that's the part that really requires a lot of thinking and it requires the input of not just subject matter experts, but people who are living and breathing the work of that organization and people who understand how hard and complex it can be to do that nonprofit's work.

Mimi Yeh

So when people are pinning their hopes on AI to fix all of that, it can help with some of it, but it depends so much on the prompts, the thought process, the inputs that are used, and then the whole human-in-the-loop experience of creating that generative and eventually stronger and more uh closer to the culture and the ethos and the values that you would like everybody in your organization to know and to build upon.

Mimi Yeh

If you want to use AI for that, it can do it, but it's gonna need a whole lot of very structured input and very clear guidance. And that's something that the human has to put in there.

Carolyn Woodard

I I love it. Thank you so much for kind of wrapping up how that human in the loop ideally works. I just want to thank you so much for your time today, Mimi. This was just a wonderful conversation, and I really appreciate being able to talk with you about it.

Mimi Yeh

Thank you, Carolyn. I appreciate it too.