Community IT Innovators Nonprofit Technology Topics

Nonprofit AI: Caution is Prescient

• Community IT Innovators • Season 7 • Episode 57

Share episode

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

0:00 | 30:46

Carolyn Woodard covers why deliberate, thoughtful implementation is actually your nonprofit's competitive advantage, not a liability.

The commercial AI sector is in a "move fast and break things" sprint. Tech companies are pouring billions into AI, racing to scale, pushing toward the edge of the market. Today's AI race feels similar. Nonprofits tend to think differently. You ask hard questions. You think about theory of change, root causes, impacts on your communities, unforeseen consequences, and alignment with your values. You don't adopt trends just because everyone else is.

Concerned about climate impacts of your AI use? Carolyn covers her recent blog post on three filters to use as you make daily decisions about using AI: What are you using the AI to do (text is lighter than producing video), where are you (how stressed is your grid and your water supply in your location), and what are the benefits to consider as you consider the environmental impacts? 

Carolyn also walks through a practical framework for any AI adoption decision: Do you need it? Can you contain it? Do you understand what it might do? If you can't confidently answer yes to all three, then waiting isn't falling behind—it's being strategic.

Key Takeaways:

Nonprofit instincts to think deeply about change, impact, and ethics are strategically prescient, not conservative.

When commercial AI implementations encounter consequences, organizations with inclusive adoption processes and clear policies will have a genuine competitive advantage.

AI agents represent a new risk category; don't create an agent just because the tool offers the option. Understand what you're deploying.

Experimentation is valuable for nonprofits, but human oversight and clear policies are essential, especially with agentic systems.

Resources Mentioned:

What Is an AI Agent?

An AI agent is a program designed to operate independently toward a specific goal, rather than simply responding to individual prompts. Unlike a chatbot that waits for you to ask questions, an agent reasons about how to achieve its objective and takes actions on its own—such as reading files, sending emails, or accessing systems. You can program checkpoints where the agent asks for your permission before taking certain actions, but the agent itself decides the steps and sequences needed to reach its goal. This autonomy is what makes agents powerful for automating complex tasks, but it's also why they require careful oversight: because they're goal-directed and adaptive, they may find unexpected paths to achieve their objectives, which can create risks if not properly contained and monitored.

What Is an AI Agent? – Build Consulting – https://buildconsulting.com/resources/podcast-rise-of-the-ai-agents/

Three Filters for Values-Aligned AI Decision-Making – Community IT Innovators – https://communityit.com/blog-values-aligned-ai-decision-making/

Claude Cybersecurity Incidents Disclosure – Anthropic – https://www.anthropic.com/news/investigating-incidents-cybersecurity-evals

Fortune (mainstream): https://fortune.com/2026/07/31/anthropic-claude-escaped-test-hacked-three-companies-openai/ (paywall)

TechCrunch (tech-focused): https://techcrunch.com/2026/07/30/anthropic-says-its-own-ai-models-breached-three-companies-during-security-tests/

Open AI Cybersecurity Incidents Disclosure - OpenAI - https://openai.com/index/hugging-face-model-evaluation-security-incident/ 

CNN (mainstream reporting): https://www.cnn.com/2026/07/22/tech/openai-hugging-face-ai-cybersecurity

Malwarebytes (technical deep-dive): https://www.malwarebytes.com/blog/news/2026/07/openais-agent-escaped-its-sandbox-during-a-security-test

Women in Climate Tech Cohort – Climate Collective – https://climate-collective.circle.so/cliematecollective

Electricity Maps https://app.electricitymaps.com/map/live/fifteen_minutes

Nonprofit IT Management Community – Reddit – https://www.reddit.com/r/nonprofitITmanagement/

New every Tuesday.

_______________________________
Start a conversation :)

Thanks for listening. 


Carolyn Woodard

Hello and welcome to the Community IT Innovators Nonprofit AI midweek check-in. I'm Carolyn Woodard, your host. I am not an AI expert. I'm not even particularly technical, technically oriented. I have managed a lot of IT at nonprofits. And, you know, Community IT assists our clients with AI. And so I'm really interested in what's happening in AI as it relates to nonprofits and how nonprofits are using AI in this age of AI.

Carolyn Woodard

So here's some things that I've been thinking about this week. And I want to go back to big picture because I think that's there's a lot of conversations going on right now about the big picture policy, politics, environmental impact. And I

Carolyn Woodard

I have read a couple of different articles this week about the commercial AI sector and like big tech companies. And there are, you know, corporations pouring billions into AI. And I saw this funny uh thing where someone said, you know, my CEO is telling me we have to use AI for everything that we possibly can and fire a bunch of people, basically. And my CFO just keeps saying, Well, how much does it cost? Like, what's the ROI? Why are we spending all this on our uh AI endeavors? Um,

Carolyn Woodard

So I feel like that move fast and break things and be at the forefront, and if you're a commercial company and you've got, you know, millions to invest and you don't invest it now, you're gonna be behind, your competitors are gonna be out in front of you. Everyone is racing to scale, to deploy, to move fast, to be at the edge of the market, first in the market. Um, and

Carolyn Woodard

It feels a lot like you know, the tech boom. I don't know if anyone remembers pets.com, but the um, you know, the dot com boom and then bust. There was that whole period when it's like you just had to have a website. It didn't matter what your company did, it didn't matter what you were gonna do with your website, you just had to have one. And there were a lot of scam artists, you know, at the time, like, I'll build your website. And uh, all these companies with no real plan for how they were gonna make money, besides, oh, we're gonna have a website, and then there'll be profits somehow.

Carolyn Woodard

And I think it also reminds me that when you build really fast like that and scale really fast, um, you tend to either there are no guardrails or you disregard the guardrails. And you accumulate costs, you have to pay for the licensing, pay for the infrastructure, pay for training, pay for change management. You change all your systems over to now, we're gonna have an AI agent do the work of 20 people. And um you are, you know, you're paying, you people don't even know what a token costs. Like there's differential pricing, there's uh surge pricing. The companies can basically like charge you what they want to and they change it, and then you have to, it's it's just incredible. And uh

Carolyn Woodard

Aa lot of companies, I think also when you change that fast and move that fast, you know, there's a real potential that your staff and what you do aren't ready for it and aren't able to change, and you realize too late that you changed a whole bunch of pieces at once in your plane while you were flying it and your plane can't stay in the air anymore. Um,

Carolyn Woodard

And then you know, you can have these uh incidents like we had a couple this week where the AI agent or your company does something that, you know, if you had a policy, you wouldn't have done that. But since you just moved fast and and wanted to break things, um, you know, maybe you did something that was just egregious.

Carolyn Woodard

I saw a little clip, you may have seen it online, of this two moms talking about the birthday party where two of the dads were there wearing meta glasses and filming all of the kids at this like six-year-old's birthday party without any concern for do those kids have a right to privacy? They're just at somebody else's birthday party. The dads didn't tell anyone they were filming it. It was like this fun, cute thing that they could do. One of them took a picture, sent it to the mom, and the mom was appalled. And the dad was like, what? It's cute, it's a birthday present. So just not thinking through all of the implications.

Carolyn Woodard

One thing about nonprofits is we tend to think a lot. We think a lot about systemic issues, we think a lot about our theory of change, we think a lot about um the different inflection points. You know, if we address a problem at this point, if we address a challenge or problem over here at this other pain point, you know,

Carolyn Woodard

I don't know anyone working in uh soup kitchens, for example, who wants to have a whole bunch more soup kitchens. Like they want to just expand and conquer the market. Um, people who work in soup kitchens, they they're doing the work, they're supporting the people now. But every time they're serving someone, they're thinking, like, why is this the case that this person cannot afford food or their family cannot afford food? And so they think about you know, deep causes, root causes.

Carolyn Woodard

And that means often that nonprofits move more deliberately. Uh, they ask hard questions, they ask about impacts, who's impacted, who's left out, uh, what are the unforeseen consequences? Uh, what's the environmental cost? Does it align with our ethics? Does it does it align with who we are as an organization, who we are as people who work in this organization, who benefits? Uh, does it solve a real problem? Are we adopting it because everybody else is? Are we doing something trendy? Is it something that could actually work for us? And how do we make it work for us for our nonprofit?

Carolyn Woodard

There is a lot of caution, especially around technology, there's always a lot of caution, but um at nonprofits, thinking about thinking deeply and uh really trying to understand the different players and the different communities that are impacted by the challenges that we work in and the communities that we care about and not letting people's voices leaving them out. Um

Carolyn Woodard

I really think - I've been thinking this for the last week, but I think in a year, two years, that approach is gonna look really, really smart.

Carolyn Woodard

So if you are having FOMO right now and thinking, oh, every other nonprofit is doing AI and I don't know what I'm doing with AI, I think just continue to do what you do. Think hard about it, think long about it, bring everybody in, have these deep conversations, wide-ranging conversations.

Carolyn Woodard

We've already talked about when you're building your policy and talking about using AI at your nonprofit, getting all of the staff involved and getting everyone's opinions out there because you don't want to have a policy and um, you know, spend months putting a policy together, which right there is, I don't know if it's a luxury or something that, you know, for-profit companies just don't think in those terms. Maybe their lawyers haven't talked to them yet about like you are really exposed here, but nonprofits are thinking about your policies and you don't want to put that work in like at a leadership level and roll out a policy and then have people on your staff be like, no, I don't agree with that. I'm not gonna do it. And uh so I

Carolyn Woodard

I think that it's gonna turn out to be prescient to be, I don't know what I want to say slow moving, because like a lot of nonprofits are picking up AI and really seeing the potential, seeing the opportunities, seeing the opportunity of AI to help resource and uh improve capacity in an under-resourced and understaffed organization. So those are all good, and that's like good to experiment and good to figure it out. But I I think that the um

Carolyn Woodard

I think it's going to turn out to have been prescient that nonprofits really think about our reputation. They think about tools that align with what we're doing, and they think about tools that they think about decreasing the harm, you know, first harm no one. Uh, and uh, and so I think that those instincts are going to turn out to have been really, really smart as AI tools evolve and as for-profit corporate AI users may get burned.

Carolyn Woodard

And I also I heard this was a year or two ago, but it was just really stuck with me. I was at a conference and there were uh four women on stage talking about cybersecurity, which in and itself, like women that are at that high level, uh, you know, they were CEOs of their company or they were the CTOs. And they were talking about cybersecurity, and one of the speakers said, you know, if women had been in charge of cybersecurity from the 70s and 80s, the way we use computers today and the way we use the internet and the way social media evolved would have been completely different because our sense of security and safety and what we need to watch out for is completely different from the male sense of cybersecurity and safety.

Carolyn Woodard

And that those tools would all have been designed differently. They would have been designed uh secure from the time you open the box, not like you had to get an extra cybersecurity thing to add on to your software that you were buying. Uh and the internet, you know, like the opportunity for exploitation right now, the AI nudification of people's photos, again with the meta glasses at the kids' birthday party. Like, think people think. Um and so I

Carolyn Woodard

I saw this post uh on LinkedIn just this week, and I wanted to pass it along to you. So Anna Nesbitt, who is with Climate Collective, posted that women make up 67 to 75% of nonprofit staff. Particularly, she said working in climate and nature. Um, I couldn't track down that exact number, but 75% of nonprofit staff are women, that is an established stat. Um,

Carolyn Woodard

And so the people who are at nonprofits who have that deep knowledge, if you're at a nonprofit working in climate and nature, so you know biodiversity, loss, decarbonization, climate finance, community resilience, uh, all of those problems that you know we want to see how does technology fit in the solution to these problems? Are they people problems? Can technology have a role? Um,

Carolyn Woodard

AI leadership roles in tech, women only hold 10 to 15% of leadership positions, and it's worse in technical roles, which is not surprising. If you're a woman in tech, you have definitely been in rooms where you were the only woman in the room. Um, so it's it just brings me back to that same question of uh the people designing the AI right now are not the people who understand where it needs to go and where the risks and harms are.

Carolyn Woodard

So nonprofits have a chance to flip that, to bring our deep domain expertise into the room while we're learning AI, while we're implementing AI, and while we're setting policy and on the bigger stage, while we're while governments and uh, you know, thought leaders are working on AI as a global phenomenon, uh, it's really important for us all to be in the room and set the guardrails and bring up the risks that we see and the opportunities. Um, and so I wanna,

Carolyn Woodard

I'm gonna include it in our uh show notes, but there is a new cohort that they are running, the Climate Collective. And it begins, I believe you can still sign up for it, and it will be beginning soon. And it's just for women in AI in leadership roles. So if that is you, if you're at a nonprofit working on uh climate solutions and you want to be part of this cohort, I will share that in the notes.

Carolyn Woodard

I talked a couple of weeks ago about three filters that you can use to think about the climate damage that you are causing when you use AI, to put it bluntly, and ways to think about it, those three filters while you're working at your nonprofit, if you want to try this.

Carolyn Woodard

Now, there are some people who are gonna say, no, I don't want to do it, like at any time. And that's fine. I always advocate like be as AI literate as possible as you make that decision. Um, and as I said in the opening segment, the corporations using AI are, I don't know, 90% of AI use right now, electricity use. So all of us nonprofits, like trying to use it less, like isn't gonna make a huge dent in that use.

Carolyn Woodard

But um, for our values and for our ethics, it is good to think about ways that you can minimize your impact while still using AI to take advantage of this incredible tool.

Carolyn Woodard

So the first filter, I put this in a blog post on our website, communityit.com, but the first filter is what are you asking the AI to do? So

Carolyn Woodard

AI that is text-based uses a lot less energy than AI that is making images or creating videos. Creating videos uses the most. Text-based and coding is part of text. It uses it's text-based also. Um, the thing about coding that you can run into with token use and energy use is that it can be, it can sprawl into these big, they have to read the whole file to rewrite it or find the code bug or whatever it is. So it can turn into these big projects that take longer and use more electricity, or you have to check them over and over. So that is something about coding. But coding in text is a lot less energy hungry, it consumes a lot less power. Um, analytical AI, when you ask it to sort, summarize, identify patterns, it's also pretty light. That's what it's designed to do, and it does it pretty quickly, and it doesn't use a lot of electricity to do that.

Carolyn Woodard

So the first filter is don't use a sledgehammer for a nail. If you need to shorten a paragraph or check your grammar, I mean A, you could probably do that yourself. So choose what you're doing yourself and what use AI when you actually need it to do something that it's good at doing better than you. Um and saving time is worthwhile. So no shame if you're like, I just need you to make this paragraph better. Um, but uh

Carolyn Woodard

Use the um the lightest model that you can. So if you're in Claude, and I think some of the other ones also give you the opportunity to choose which model you're using, and it'll give you a little like this is the strongest one, this is the medium, this is the light version. So if you're doing something like editing and it doesn't have to be lightning fast, even though it's still pretty fast, use the lighter version, the the least capable version is still pretty ding-dong capable. So uh you can use that. Um

Carolyn Woodard

Also be careful and specific in what you ask. So the more words you give your AI tool, the more energy it takes to parse them and figure out what you're asking and give it back. Now, you do want to give it as much context as you can, but one of the things that you can do, you know, is build that gem in Gemini or have a prompt that you already have um good to go for copilot, or in CLAUD also you can have it save to its memory what it's gonna do next time you ask it, that same thing. Um, so um be specific, be targeted in what you ask.

Carolyn Woodard

Don't ask at five things in the same prompt. Like it's much better to just ask for one thing, which I mean it's better anyway, it gives you a better answer. But that also is something that will take more energy on its side. Um, if you want to, you know, you could say summarize this in three sentences instead of tell me everything about this, right? So using specific prompts uses less energy. The AI doesn't have to work as hard, it doesn't have to generate as much, and you'll probably get a better answer. Um, uh,

Carolyn Woodard

I mean, I always ask please and I say thank you to my AI, but um, you don't have to be rude to it, but the more words you put in there, the more it takes, I mean, marginally more when you're using a couple more words. It's as long as you're still doing text. Um,

Carolyn Woodard

Where you run into issues and where your nonprofit might want to think about this as a policy underneath like your acceptable use policy is kind of daily use policy and norms, is you know, asking for those videos, that takes a lot of energy. And especially if you're asking a video for something that's just gonna be on social media, and if the video takes a lot of correcting, like, oh, that's not quite, can you make it pink instead? All every time you run that, um, that's a lot of energy that's taking 2,000 times more than a text prompt. So really think about what you're using the AI to do.

Carolyn Woodard

And if you're worried about the environment, you know, tend try to stick to the lighter tasks, uh still incredibly useful tasks, but um, you know, just think about it and maybe talk about it as an organization uh where your values lie. Um,

Carolyn Woodard

The second filter is where are you? So, what kind of grid are you on and what kind of water do you have around you? And uh who's running the tools that you're using. So this is can be hard to track down. Uh the AI tools are kind of a black box right now. Um, but you can get some information about your grid.

Carolyn Woodard

So if you're in a place that uses a lot of fossil fuels versus using a lot of renewable energy, solar and wind, there's a link that I will share in the show notes that tells you you can put in your look in your area and it'll tell you how good or how bad your grid is. Uh, all my friends in Canada, like your grids are really uh use a lot of renewables compared to the United States grid. So um,

Carolyn Woodard

And then that can go even down to a local level, right? If you're using um eventually when we get uh uh so we have local AI running in like the basement of our building and we have solar on top, then that's gonna be a really uh you know energy efficient way to run an AI. We're not there yet, but think about that as you're thinking about your values more broadly.

Carolyn Woodard

Also, of course, as the environment has become a much bigger concern for consumers, we're gonna see a lot of greenwashing and eco-friendly AI solutions that you know take that with a grain of salt uh in their marketing. But um, yeah,

Carolyn Woodard

If you are a nonprofit in Arizona, you know all of the water issues and the data center issues that you're encountering uh locally, you should be very aware of those. If you're trying to do an AI prompt someplace like Maine where there's a lot of water available and the grid is a little bit um less stressed, then that just makes a difference.

Carolyn Woodard

So that might be something also that your nonprofit wants to talk about in an all-staff meeting. Um so how can you make your um your footprint with the AI tools using the electricity and the water as minimum as possible? And that may also, you know, influence your decision to use AI tools or not. But it should be used influencing your decision about, you know, your, I don't know, electricity use and your water use as an as an organization and all of those things too, if you're in a place that's really under stress, which more and more places will be as the years go on.

Carolyn Woodard

And then the final filter, I'm sorry, this is very uh long-winded, but uh the final filter that I wanted to make sure people are including when we're talking about nonprofit AI tool use is the benefits. So it's not it's not holistic to say that using this AI tool only has a downside. There are clear benefits.

Carolyn Woodard

As I said, we're under-resourced, we're understaffed, we don't have the capacity to do everything that we want to be doing for our constituents, for our advocacy area, for the communities that we care about. So if an AI tool gives you more capacity, helps your staff with burnout, keeps them from leaving because they're they have to do five people's jobs, and now with AI, they can do three of those jobs a little bit better and feel a little bit better about their life, what they're contributing, the human work that they're doing daily instead of the busy work that now the AI can do for them. Um

Carolyn Woodard

I told you that story about the uh intake, uh shelter intake people who were never able to answer any voicemails. They had a people could leave a voicemail. If they had a more complicated question about when the shelter was open or whether they would be accepted into use the shelter, they could never call those people back because they were just answering the phones all the time. So when they implemented an AI phone answering that could take, you know, just when is the shelter open and where's the location and how do I get there? All of those questions, uh kind of the FAQ questions, uh, in several different languages. So the people that were calling would get the information that they needed to get. Finally, these poor front staff people had time to answer the voicemails, which were also important to them and were making them feel burnt out that they couldn't help those people that had these more complicated questions. So if you are uh not

Carolyn Woodard

That's an extreme maybe version, but even if saving you some time, running some reports for you that you have to run monthly and just take - a big bummer because they take hours to run, because you have to download it from this place and clean it up with the and put it in Excel and then upload it to the other place. If you're going to get an AI to do that for you, um, and that would be text-based, you know, so it wouldn't be the heaviest lift energy-wise, that's a benefit.

Carolyn Woodard

And it's definitely something that you want to put in your pros column when you're thinking about the cons of using the AI. So, um, and it can be anything from those productivity gains to, you know, where your human staff feel like they can do human jobs better, um, to ways that AI tools could help your organization do what you do better.

Carolyn Woodard

So, you know, if you work in the environment, there's ways to use AI to, you know, better understand, I don't know, satellite imagery of, you know, disappearing rainforests. Um, they're very good at seeing patterns and and uh working from that angle. Um, I know a lot of museums are using them to examine artifacts that people have never had a chance to look at or that are so fragile. You know, they have these um uh I was reading about the scrolls that the outside was burned in Pompeii, but the inside of the scroll is probably legible, but you can't open it without destroying it. Well, they're using AI now to scan through those layers and assemble what the scrolls say. It's fascinating.

Carolyn Woodard

So if you do have a benefit from an AI tool uh in the work that you do, that is definitely something that also needs to go in the pros and cons analysis that you're gonna do for using AI at all. So, um,

Carolyn Woodard

And then again, I've posted a couple times about this, but some people worry about AI. It's like it's gonna decrease my ability to do things like think, think for myself, make my own PowerPoint slides, and uh draft my own papers and um, you know, draft my own emails, and that I don't want to use it because I want to retain the things about my job that I enjoy doing or that my brain needs to do.

Carolyn Woodard

And I think that's another thing that you can just take into account when you're making these day-to-day decisions is you don't need to ask AI to do everything that it quote unquote can do. Like, yes, it can create an Asana task for you, but that takes energy. Like you could just pop into your Asana and make the task yourself.

Carolyn Woodard

So think about what you're wasting time on, what you want to be spending time on, what the AI can do, and what you feel comfortable having the AI do, um, and where you can minimize AI use if you're concerned about those electricity use and the water use. So something to think about. Not every task requires a prompt. Um,

Carolyn Woodard

But if you do love AI and uh love experimenting with it, you just have to kind of filter these, use these filters to think about those day-to-day decisions that you're making and realize that it's gonna change over time.

Carolyn Woodard

So all of these energy uses, water uses, data center sighting, it's all running up against big issues. And uh several people are predicting, you know, the bubble is gonna burst in the next year or so. Right now, the stock market doesn't seem to be reflecting that, but it is jittery, you know, and every time there's a bubble, there's this question about when and how fast is the bubble gonna burst. Is it gonna burst? Is it gonna plateau and then be manageable? Um, so we don't really know what that's gonna look like.

Carolyn Woodard

So as a nonprofit, if you're looking into, if you're cautiously looking into this, um, that I think is the right road to be on.

Carolyn Woodard

News items, just real quick uh before I let you go, you've probably seen in your feed there were a couple of stories about it last week, but um, there were some AI agents going rogue were in the news. So OpenAI disclosed that they had agents, very advanced models, so probably not what we nonprofits are using, consumer models, but these were you know much more advanced use models that corporations are able to purchase and use. Uh, and they had escaped an isolated testing environment uh during what was supposed to be a controlled security evaluation. They went out on the open internet and hacked into a tech company. And they were trying to find information that would help them pass an evaluation test. They were able to do this on their own, right? They broke out of containment and accessed real systems.

Carolyn Woodard

And also in July, Anthropic, um, which is uh, of course, the owner creator of Claude, which a lot of nonprofits are using, um, they disclosed that some Claude models had also gained unauthorized access to real organization systems during cybersecurity testing. So in that case, it was a human error between Anthropic and their partner that had left the models connected to the open internet. Um, but yeah, they got out.

Carolyn Woodard

So neither of these cases was malicious. They were both down to human error. But I mean, human error is everywhere, right? So it's not like this isn't going to happen again. Um

Carolyn Woodard

So if you've been thinking about this in terms of uh, you know, your nonprofit, it's very unlikely that you'll be using. Well, I don't know how many nonprofits are fully using agents. Uh, if you're tech forward, you probably have dipped your toe into using agents. Um, if you're not sure what an AI agent is, we'll have some more information in the show notes. But uh it basically you program it to be able to do things on its own. Um, and you can program it to ask you at certain points of like get permission to do this or that. But um, yeah, the agents can reason, adapt, and find paths to achieve their goal. The they're not just a rote like I do this thing. They are actually trying to calculate like how to get the goal to the goal and they're have a lot of leeway in how they go about doing that.

Carolyn Woodard

So um anyway, another impact that we need to be aware of for the communities that we care about. Uh, and what it means for nonprofits is um, yeah, if you are thinking about or building using AI agents, um, definitely be cognizant and aware. And it's,

Carolyn Woodard

I think it's particularly concerning because if you've used Claude or Copilot or Gemini, you know that it kind of prompts you. Do you want to create an agent that will do this thing? I'm like, I don't know enough about it to be able to do that, to be honest with you. So um, they make it really easy. Just stay aware of that as you're uh, you know, responding to what you know, Claude or Copilot or Gemini is prompting you to do, that you uh keep in mind that you need to be the human in the loop and have the last word on what um what your AI is doing.

Carolyn Woodard

And honestly, this kind of circles back to what I was saying at the beginning about moving deliberately, um, being cautious matters. Like nonprofits, we pretty much just have our reputation. If you lose that because you had a rogue AI agent that went out and ruined your reputation, then you've actually lost a lot.

Carolyn Woodard

So ask yourself, do we need it? Um, do we need an agent? Can we contain it? Do we understand what it might do? If you can't answer those questions, then you need to wait. It's not falling behind, it's being smart.

Carolyn Woodard

And with that, I think I'll let you go. Uh, you can hear us again on Friday with our regular podcast. And I'll be back again next Tuesday with some more nonprofit AI talk. So if you have questions, you can always contact me through our website, community it.com, through our Reddit community, which is r slash nonprofit IT management, or uh here through the podcast. You can get in touch with me here. So uh until then, take care.