Community IT Innovators Nonprofit Technology Topics

Nonprofit AI Governance Tips with Nura Aboki pt 1

• Community IT Innovators • Season 7 • Episode 41

Share episode

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

0:00 | 32:46

In the first part of this two-part conversation, Carolyn Woodard and Nura Aboki, Senior Consultant at Community IT Innovators, dig into what it actually looks like to implement AI at a nonprofit - not in theory, but on the ground. With 80% of nonprofits using AI without any governance policies in place, unmanaged adoption is already happening. This episode helps nonprofit leaders understand where to start.

Nura draws on real client experiences to walk through two case studies: a nonprofit that had to pause its AI rollout to answer a fundamental "why are we doing this?" and an environmental organization that wrestled with whether using AI conflicted with its mission. Both examples illustrate why values alignment and change management have to come before any tool selection.

Carolyn and Nura cover:

  • Why starting with a clear "why" before selecting any tool is the single most important step in AI adoption.
  • What AI literacy means for nonprofit staff and where to find free and low-cost training options.
  • A step-by-step framework for intentional AI implementation: communication, piloting, due diligence, and layered training.
  • The risks nonprofits need to plan for, including shadow AI, vendor churn, data privacy, and the legal reality that humans are accountable for AI errors.
  • Why appointing an internal AI champion matters even at - especially at - small organizations.

Resources Mentioned

_______________________________
Start a conversation :)

Thanks for listening. 


Community IT Intro

Thank you for joining Community IT for this podcast, part one. Subscribe wherever you listen to podcasts and leave us a rating to help others find this leadership resource for nonprofits. Listen for part two in your podcast feed.

Carolyn Woodard

Welcome to the Community IT Innovators webinar on nonprofit AI governance tips. We've been doing a lot of webinars and blog posts around AI, and we have a new nonprofit AI podcast that comes out on Tuesdays covering news and resources for nonprofits.

Carolyn Woodard

And a recent study found that 80% of nonprofits are using AI without any governance policies at all. So that's unmanaged adoption and it leads to lots of risks, leads to everyone in your nonprofit heading off in different directions when they're using the AI. So I'm really excited today to talk to our senior consultant, Nura Aboki, who has been assisting our clients in implementing AI for over a year now. A lot of the content about AI at nonprofits can be kind of theoretical or strategic, but today we're going to get a chance to hear from him about real life nonprofits and the decisions they're making around AI tools and adoption. So

Carolyn Woodard

My name is Carolyn Woodard. I'm the outreach director for community IT. I'll be the moderator today.

Carolyn Woodard

Very happy to hear from our expert, but first I want to go over our learning objectives. So today we want to focus on these themes. What do you need to know about AI implementation? What is AI literacy? What is AI governance? And what are good models to follow for decision making and policy making, what are AI risks, and some emerging best practices from some case studies.

Carolyn Woodard

And now I'd like to let Nura introduce himself.

Nura Aboki

Thank you so much, Carolyn. Hello, everyone. I'm Nura Aboki. I'm excited to be here. As Carolyn mentioned, I've been helping nonprofits explore and adopt AI tools over the past year or so. I really look forward to sharing practical stories and tips from those experiences. And I'm eager to answer your questions as we go through today.

Carolyn Woodard

Yeah, we're going to try and leave some time for questions. I have to tell you at the outset, we were just going over our slides and we packed a lot in. So we may have to go over some of the slides fairly quickly. They will all be on our website. We'll have the transcript. I'll include all of the links that we share and resources on our website as well, communityIT.com. So don't worry about it if something kind of flies by. But do if you have questions at any time, go ahead and put them in the chat or in the QA.

Carolyn Woodard

Before we get started, if you're not familiar with Community IT, I'm going to tell you a little bit more about us. We're a 100% employee-owned managed services provider. We provide outsourced IT support exclusively to nonprofit organizations. Our mission is to help nonprofits accomplish their missions through the effective use of technology. We're big fans of what well-managed IT can do for your nonprofit.

Carolyn Woodard

We serve nonprofits across the United States, and we've been doing this for 25 years. We are technology experts and are consistently given an MSP 501 recognition for being a top MSP, which is an honor we received again in 2025. And we believe we're the only MSP on that list serving nonprofits exclusively.

Carolyn Woodard

I want to remind everyone that for these presentations, Community IT is vendor agnostic. We only make recommendations to our clients only based on their specific business needs. And we never try to get a client into a product because we get an incentive from that or a benefit. We do consider ourselves a best of breed IT provider. It's our job to know the landscape, what tools are available, reputable, and widely used. And we make recommendations on that basis for our clients based on their business needs, priorities, and budget.

Carolyn Woodard

As I said, we're going to leave as much time as we can for QA, but I don't expect it to be very much time. You can submit your questions through the chat or QA feature at any time today. And of course, you can always join us there and ask questions and we check that so we will be able to answer them later, too.

Carolyn Woodard

A little bit more about us. Our mission is to create value for the nonprofit sector through well-managed IT. And we also identify four key values as employee owners that define our company: trust, knowledge, service, and balance. We seek to always treat people with respect and fairness, to empower our staff, clients, and sector to understand and use technology effectively, to be helpful with our talents, and we recognize that the health of our communities is vital to our well-being and that work is only a part of our lives.

Carolyn Woodard

So now I'm going to launch our first poll. The question is: do you have AI policies? And the answers that you can choose are, I don't think so. There's no shame. You've come to the right place. Second answer is we are in the process of creating policies. Third answer is yes. Our organization has created an AI acceptable use policy. The fourth answer is yes. And our staff understand our policy, you know, are using it every day. And the fifth answer is not really applicable. So maybe you've joined our webinar and you aren't at a nonprofit. Uh, you're just here to learn more about it. So it doesn't really apply. And Nura, can you read the results?

Nura Aboki

Yeah, uh quite interesting here. So yes, I'm really impressed by what I'm seeing. How about 21% uh said we don't have an IT uh policy? I don't think so. And you know, this is we expect that several organizations, nonprofit organizations would not have one.

Nura Aboki

But what I'm really impressed by is a lot of uh 53% actually said they are creating one, they're in the process. So that's quite impressive uh to see that. And hopefully some of the tips we share today uh would help in that endeavor. Uh, and then we have the those that do have about 15% already on AI acceptable use policy, and then there's about 6% that does have acceptable policy AI, and staff know about it.

Nura Aboki

So quite impressive. That's ideal. But don't worry, wherever you are on this spectrum, uh our content should uh be useful to you. So we're going to we're going to start by laying out some of the foundational uh principles uh about AI and nonprofits, and then we should be able to be able to give you additional contact, whether you are a beginner or already advanced and have staff know about this.

Carolyn Woodard

Yeah, that was really interesting. And congratulations, as Nura said, to everyone who is working on it. And uh, if you haven't started yet, we're hoping that this will give you some place to start from. Uh, but thanks for coming because clearly it's important. Uh so

Carolyn Woodard

This slide actually we shared in January in our tech technology for nonprofits roundtable. And I just wanted to share it here because that feels like you know 100 years ago in AI year terms. But um, these are some of our talking points and kind of what we stand by that you need to match your tools to your needs, beware of freemium tools that you don't aren't paying for a license, so you don't have a relationship with that, a business relationship with that vendor, making sure the human is the last editor. You'll hear that called human in the loop often when they're talking about AI, having a policy, keeping training and talking about it. The AI is being added to tools that you already use. So you need to take that into account. And again, just taking training seriously and upskilling yourself and your staff. So

Carolyn Woodard

I felt like this slide, you know, kind of illustrates the tools and AI is changing really quickly. Um, but there's some emerging truths that are happening, which is that the underlying principles of how you use AI are not changing as rapidly as the tools are changing. So having a strong IT foundation allows you to adopt new tools well.

Carolyn Woodard

AI is a new tool. It's very disruptive, it's changing our work and our sector, but it is a tool. And our past experiences with tool transformations can help you in this instance, too. So, for example, when a lot of nonprofits moved to the cloud, that was a similar time where everything was changing. Having a good foundation, having policies, and really being transparent about it are all very important when you have a new tool that's like coming in so quickly. Um,

Carolyn Woodard

I want to share a couple of resources with you as well, just quickly. I will share the uh link in the chat with you. This is from a Department of Labor publication about AI literacy and how it is changing the workforce. I found it very helpful. It's a short PDF document. You can find it online. It has probably four or five pages about each of these points. So I'm not going to go into deep on any of them now. Just wanted to share this graphic with you. They have a free download. They have excellent ideas on training, hiring AI literacy in general. So

Carolyn Woodard

They start with what AI literacy entails. So understanding what AI is, exploring the use, directing the AI effectively, evaluating AI outputs, and using it responsibly, taking into account ethics. And then on the flip side, also part of this document, they have their effective delivery principles of AI literacy. So basically, training. How do you go through training with your staff? And they have these different touch points. And as I said, in the document itself, they have a bunch more of uh you know details on how to do these seven different points.

Carolyn Woodard

I'm also going to share with you in the chat some uh resources around AI literacy that are specific to nonprofits. So TechSoup has some courses, Nten has a certification, there's the Human Stack, NetHope. Um, you can also search for on Microsoft, Google, and Anthropic for free AI training for nonprofits. Um, if you did an update on a tool or a platform you already use, and you notice there's a new AI tool, like a little icon or a helper, a chat bot, uh, usually you can access that tool's knowledge base and learn more about how to use the AI within the tool. So, for example, in your CRM, Zoom, MailChimp, Google Analytics, they're all introducing AI tools that can help you. But that's something where just look in the tool and it'll give you more information about how to use the AI. Um, so

Carolyn Woodard

I wanted to also share this link to an article that we put out from our president and COO, Johanny Torco, uh, where she just did a really good analysis, I think, of a mission-aligned AI adoption model. It's a free download. You can get it on our site, and I shared it in the chat with you. It'll be in the transcript. Um, I think we can say at this point, 100% of nonprofits have staff using AI tools. But as I said earlier, uh there's a recent survey that came out with TechSoup and Candid where they found that only 80% of nonprofits uh have an AI policy. So I'll share that article with you as well if you haven't seen it already.

Carolyn Woodard

And now I want to turn it over to Nora for our first case study. So, Nora, you talked about um some organizations that you worked with with AI implementation. So, can you give us a little bit of how they went about it?

Nura Aboki

Yeah, thank you, Carolyn. Uh, I'd like to share two real world uh scenarios here. You know, I've encountered that you know, what nonprofit organizations uh need uh to consider at the start of AI adoption has to do with uh looking at AI policy and governance.

Nura Aboki

So one scenario was a nonprofit executive suddenly realized staff were using uh AI tools without any policy, which is kind of commonplace uh these days, and said we need a policy now. Uh people were already experimenting with AI and there was no guidance. So when leadership stepped in uh with the urgent mandate, staff were uneasy. Uh those uh who hadn't used AI were asking, why do we need this tool? What's the plan? Why should we trust these tools? Uh

Nura Aboki

The leadership quickly saw that there would be a need to provide clear strategic rationale or guidelines, essentially just answering the staff's question of why uh before and AI rollout would succeed. So uh they took a pause uh to define uh how uh AI aligned with their goals and what guardrails uh would be uh used uh to keep them safe, and only then to resume into implementation.

Nura Aboki

So this asking of why, even if leadership says you need a policy, it's very important for leaders, for uh uh stakeholders and organizations to understand with that clarity why do we need to use these tools and why do we need a policy to guard to guard real source? Because uh if that is not clearly understood, then you behold have buy-in and we'll have a direction and it says we would have a better strategy on how to use uh AI tools. Uh

Nura Aboki

I would want to add just a second uh example here. Uh sort of a nonprofit organization whose mission is environmentally environmental sustainability, um, were initially hesitant about AI. So you can think of what we've probably read or what we heard, uh, how AI is using up a lot of energy, uh, a lot of resources in our environment. So staff were worried. Look, our mission is to protect the environment. So why are we considering using AI? Because as we know, of course, we read about it's using a lot of resources, competent resources, which will kind of contradict their uh mission. So um

Nura Aboki

There was that internal debate. Uh can we justify using AI if our mission is not is to protect the environment? Ultimately, they reached a consensus that AI could help them in a way, and many ways that justified the uh environmental cost. For example, automating tasks uh so staff could force focus more on mission critical work, but they also committed uh to using uh greener AI options uh to offset there uh where possible. Uh but

Nura Aboki

This open values and discussion uh was essential because uh otherwise staff would just feel a sense of guilt. You know, so once everybody was comfortable and understood what uh the reason, there was that mission alignment uh of why AI, how AI could actually be used responsibly. Um, and they proceeded uh to set out policy and then training staff on how to use AI. You know, uh before just the key takeaway there is before implementing AI uh at your nonprofit organization, you want to address that fundamental why and how questions internally

Carolyn Woodard

It's a perfect segue into our second poll, which is the biggest barriers to AI governance. Um I'm gonna go ahead and launch that.

Carolyn Woodard

And the question is what is the biggest barrier at your nonprofit to AI governance? So your options are one, leadership challenges, and uh that could cover a broad range of things from the leader that you said, Nura, who you know was really excited about it and needed to make sure that they weren't going too fast for the rest of the staff, or you know, leaders that might be very reluctant about it, but they have staff that are really excited about it. So challenges and leadership and even sometimes just finding the time. Uh,

Carolyn Woodard

Number two is don't know where to start, right? Is a barrier of finding where where the information and like how to get started. Uh,

Carolyn Woodard

Three is started but stalled because we lack time and capacity. You know, what nonprofit has plenty of time and extra capacity to do anything. So from trying to prioritize this along with all of your other priorities, it can get lost. Uh,

Carolyn Woodard

Number four is we're using so many AI tools already and we haven't needed governance so far. So a little tongue-in-cheek there, because we already started off saying like everyone needs governance, but it can you can feel like you're in that situation of we all just, you know, this team uses this one, I use this one personally. We don't really need to have a full governance for our organization. So we're gonna talk through that one. But again, no shame if that's where you're at, just let us know. Uh,

Carolyn Woodard

Number five is something else, another different barrier. So if you want to put that in chat, if you feel comfortable sharing it with us, we would love to hear that. And then the other one is uh again, just not applicable. All right, so Nura, can you see this also?

Nura Aboki

Yeah, well, also here I can see that uh that there's some uh leadership uh challenges, about 3% say that. Uh that's one barrier. Uh but 36 uh percent here uh say, well, we don't know where to start. You know, they aren't sure of where to start. We have about 33% who said we've started but stalled uh because we lack time or capacity. So we totally understand, especially with no small nonprofit organizations, capacity can be a challenge, but we'll talk about that uh shortly. And then yeah, the others uh not applicable. Uh there wasn't any results for we're using uh so many tools already, and we haven't uh gotten the need for uh governance so far.

Carolyn Woodard

So it can happen, like I said, but um congratulations everyone here. It makes sense that you're in a a webinar about governance because you think governance is important. Um

Carolyn Woodard

There were some really interesting things that came in under other, so I'm just gonna quickly um shout out them thank you for using that feature. Uh,

Carolyn Woodard

One of the biggest problems is that the genie is out of the bottle, and controlling what tools people are using is proving challenging, the challenge of shadow AI. Um,

Carolyn Woodard

Another other is overwhelmed with AI intrusion into everything, paired with knowing there are devastating consequences for communities and the planet, and good information about the real consequences are hard to come by. I hear you, that is for sure. Um, other most of our staff are ethically opposed to AI and don't want to touch it.

Carolyn Woodard

A few people are interested in specific tools and instances. Definitely hear you. You've come to the right place. We're seeing that a lot. Um, another is disparities in access to free versus paid AI tools across staff. Definitely, it's hard to say, oh, we should budget for licenses for everyone. When you're like, oh, it's free, I can just go out here and do it. Um, and

Carolyn Woodard

Another one is that among other challenges, I think a big barrier is the reality that AI is impacting our communities so significantly, both the environmental impact and on our community with data centers threatening our desert region and resources, it's difficult to navigate a use policy. So much goes against our core values. So we definitely hear you. That is something we're hearing a lot from nonprofits. So you're not alone in feeling that. Um, all right, I'm gonna go ahead and stop sharing.

Carolyn Woodard

Thank you, everyone who shared with us. I really appreciate it. Um, and we're gonna move on and Nura, you have some more that you want to talk about in the many organizations that you've worked with over the past year. Some clear best practices are starting to emerge. Um, so I'd love to hear more about it.

Nura Aboki

Yeah, so now we'll start with some best practices for implementing AI. Uh, you know, these uh steps will address many of the barriers uh that that you've uh identified. So

Nura Aboki

The first one there that you see is starting with a clear communication and uh change management uh plan. Uh before anything, uh get your team on the same page because it's very important you need to explain your exploring AI and uh how it could help your mission. So that way those questions around why, why will be uh clearly communicated. So address the fears upfront that people have uh so everyone is at least open-minded about uh AI.

Nura Aboki

Second one is uh taking that uh structured intentional approach. Uh treating AI adoption like any other major project because this thing is really making a difference. So we want to take it seriously, we want to have a uh methodology and approach that is intentional. You want to bring in keys key in uh key stakeholders early and plan the phases and basically do not just uh flip a switch overnight because that's what ends up happening. Uh, you end up with a shadow AI, as someone uh mentioned earlier. So uh quick adoption might be tempting, but intentional planning uh prevents problems uh down the line.

Nura Aboki

Now the third one here really talks about appointing an AI owner, uh identifying someone who can be a champion or who's willing to be a champion and manage the AI tool in. Internally is important, but you may argue we're a small nonprofit organization. We may not certainly have someone that can be a specialist there. But AI is widely used. And if you understand how important it is for your mission, then setting aside or identifying an owner will be quite useful for your organization. It might be one person or any staff that is passionate and knowledgeable about AI or is willing to learn about AI. And that they'll be the kind of go-to person to your organization and how these tools work and its settings and preferences, and you have ready human answers ready, not having to go to Google and get a curated answer, but someone that has experienced it, that you trust, will be able to provide guidance. So appointing the owner is highly valuable.

Nura Aboki

And then we move on to the fourth point there in terms of best practices, just deciding on metrics, guardrails, and running a pilot. Defining what success looks like, uh, whether it's improving accuracy, whether you're going to save uh staff time and setting boundaries, like do not feed uh do not PII into the AI, you know, is very important there. Uh, then you want to run a pilot project. Uh, let a small group of uh staff use the tool on the setting guide rails that you've defined and then get feedback rather than open it up to the entire company or organization. You want to start small with that pilot project. Feedback is important to kind of iterate and make sure the experiences that people ultimately end up with is something that your organization is gonna have uh value uh around.

Nura Aboki

Then uh the fifth one is definitely doing your due diligence. Uh while this pilot runs, uh evaluate multiple tools, options uh with transparent criteria because the tools are evolving in the past. Uh when it started when we started seeing hearing about generative AI, it was one tool, big popular tool, for AI. But now there are other tools today, it may be cloud or another tomorrow. So definitely you want to evaluate the variety out there and consider feature cost, kind of uh the data security that is around the tool, and also that vendor reliability. What if you use a tool that it's uh you end up with a tool that a vendor closes? That could be a risk there. Uh so

Nura Aboki

Involve leadership and weighing those options and risks, and then when you pick a tool, document. Document why, and it helps everyone kind of trust the choice that has been made. And may not you may not necessarily land on one tool, it could be a variety of tools that are useful uh to your organization.

Nura Aboki

So then you want to move on to after designing a tool, providing basic training for all staff of the chosen tool. Uh, once you have that tool, training your uh staff on how to use it, or at least how staff will be affected by it is very useful and beneficial. So that will cover both how to use it and your rules uh for using it. That is the policy that you set around the use of AI. Uh, then give people time to practice solo to encourage encourage them to experiment within uh safe bounds. Because uh that those guide rails that we talked about, the policy would have those guardrails in writing so people know their limits on how to use the tool. Uh and then uh there uh

Nura Aboki

As you see lastly, there we we are seeing another practice that we are seeing encouraging our nonprofit organizations to do is to have a second training. It's not just one training that you know the basics, but you want to look at how you can harness the potential of AI by focusing on collaboration and sharing lessons. So after folks have some experiences, uh, they want to bring it back uh together uh for a deeper session on using AI tools within a team workflow or the organizational workflow. Uh

Nura Aboki

This is where you'll see real adoption and the biggest benefits uh when people learn from each other's use cases and integrate AI into everyday processes. So I would say by following uh the steps, uh communication, planning, piloting, due diligence, and training, uh you weave governance into, and you weave governance into this actual implementation. So it ensures everyone is involved, risks are managed, and your AI adoption is aligned with your mission from the start.

Carolyn Woodard

I want to uh briefly acknowledge this great conversation that's going on over in chat around the ethics of using AI at all. So great comments.

Carolyn Woodard

I think there's an emerging um kind of philanthropy/slash nonprofit um, some consortiums are emerging around using AI for for the good of you know the planet and the people uh versus a for-profit bent of AI right now, and that's something that philanthropy can really speak to. Um that's a kind of a separate conversation, but I just wanted to acknowledge that, and there are some resources out there, so I'm gonna see if I can pull some up and share those as well. Um, but

Carolyn Woodard

I wanted to move on a little bit to what is AI governance? So just to make sure that we're kind of level setting here, um, some of the great things that Nara just talked about. Um,

Carolyn Woodard

So all of these questions, right, of who decides what your AI policy is at your organization? Is it a leader? Is it an all-staff decision? Those sorts of questions. Who decides what tools get used? You know, you may want to have something that's uh very specific, like we only use these tools, those are the only approved ones, or you might want something more flexible for different teams that have different tasks that they're trying to do with these different tools. So you might have more around principles. So this is how we choose a tool.

Carolyn Woodard

So who purchases or owns the tools? Who decides on AI AI access to files? How do they monitor risks? Who is responsible when something goes wrong? How does your policy stay current? You know, how often do you revisit it? How does your staff know what the policy is? We had that in the past uh poll about we have a policy and staff are using it and know where to find it. So um,

Carolyn Woodard

These are not, you know, like this is how you do it. This is more questions that you need to take back to your organization and sit with and talk about like how do you make these decisions? Um,

Carolyn Woodard

And along with those decisions that kind of hints at some of these risks that um AI risks that nonprofits should know about, um, your data privacy, what is shared, what's accessible. When you said earlier it's hard to get people using a paid license or justify having a paid license when there's a free version, but that free version carries a lot of risks with it when you're you know uploading documents to it or making queries and prompts. Um,

Carolyn Woodard

Cybersecurity vulnerabilities that may increase because you're using AI or because cyber, you know, hackers are using AI and they have more access to what you shouldn't be doing. Um shadow AI, we talked about that earlier. Staff are just using a whole bunch of different tools, and all of that brings risk to it.

Carolyn Woodard

Bias and equity issues. So I think this is getting at a little bit what we're speaking about in the chat around how which tools you're using, how they were trained. Uh, are they are they harming communities that you care about? Are they enabling communities that you care about? And those questions that really speak to your values as a nonprofit. Uh,

Carolyn Woodard

Vendor commitment and tool churn. So, what what is your investment? How would you migrate if you change your mind? Like we thought we were using gonna use this tool, now we're deciding we're using this other tool. What if that vendor goes out of business? We're in the middle of a bubble right now. So there's a lot of uh consolidation, there's a lot of market forces going on. So you might commit to a tool and six months later it goes belly up. Um,

Carolyn Woodard

Staff confusion and distrust, staff that have different ideas about what your AI policy should be or what tools you should be using. So people have mentioned that as well in the in the uh chat and in the the responses to those polls. Uh, your reputational risk from AI-generated errors, uh,

Carolyn Woodard

Legal risks, right? Um, we say the human in the loop, but um, someone referred to it as the human in the jail because the AI isn't gonna go to jail if you do something that's not allowed with your AI. The human who approved it or who used it is gonna be the one that's responsible. Um, so

Carolyn Woodard

Nura, I wanted to just ask you quickly about that vendor commitment issue. Can you talk a little bit more about that and how to think about it?

Nura Aboki

Yes, absolutely. I recently saw a nonprofit organization's title one tool, uh, say Microsoft 365 uh Copilot, and then decided uh they like ChatGPT better, you know, and then they switch again to a newer model like Cloud, uh, which is better, which better met their needs. So this kind of tool churn uh is happening a lot. Uh it means that you, as a non-profit organization, will want to plan for the possibility of switching tools or the vendor changing their product uh and consider the training and change management needed uh each time you switch. So staff aren't confused. Uh, you know, it's important to track these changes that are happening and decisions are made effectively. Otherwise, you risk folks or your staff uh going rogue with their own preferred tools. So more Shadow IT or Shadow AI as it's called.

Community IT Intro

Thank you for joining Community IT for this podcast, part one. Subscribe wherever you listen to podcasts and leave us a rating to help others find this leadership resource for nonprofits. Listen for part two in your podcast feed.