Subscribe
Share
Share
Embed
Welcome to CharityVillage Connects – a series that highlights topics vital to the nonprofit sector in Canada. CharityVillage is a resource to over 170,000 charitable and nonprofit organizations in Canada. This series, hosted by President Mary Barroll, will provide in-depth conversations with experts in the nonprofit sector. We’ll examine diversity, equity and inclusion, innovations in fundraising, the gap in female representation in leadership and many other subjects crucial to the growth and development of charities throughout Canada.
Responsible AI Adoption for Nonprofits
Mary Barroll: Hi, I’m Mary Barroll. In this episode of CharityVillage Connects, we take a look at how AI is changing the way nonprofits work, plan, and support their teams and we’ll examine what responsible AI adoption looks like.
But first, here’s a word from our partner.
Sponsor Message: Today’s podcast is brought to you by WUSC. For more than 50 years, WUSC has been working alongside communities around the world to drive positive education and economic outcomes for young people.
Now, Canadians have the chance to join us by volunteering internationally. As a WUSC volunteer, you’ll collaborate with local organizations, share your experience, and help co-create initiatives that expand economic opportunities.
Explore volunteer roles today at volunteer dot W-U-S-C dot C-A.
SFX: Buzzing sound
Mary Barroll: Welcome to CharityVillage Connects. I’m your host Mary Barroll.
SFX: Hummingbird flying and tone
That’s the sound of a Hummingbird pollinating our world and making it a better place. The Hummingbird is CharityVillage’s logo because we strive – like the industrious Hummingbird – to make connections across the nonprofit sector and help make positive change.
We’ll offer insight that will help you make sense of your life as a nonprofit professional, make connections to help navigate challenges and support your organization to deliver on its mission.
Mary Barroll: Many nonprofit leaders want to ensure that the sector won’t be left behind, when it comes to the new AI tools already revolutionizing our world and our work. In this episode of CharityVillage Connects, we're asking leaders in the sector how AI can be used by nonprofit organizations to streamline workflow and add efficiencies. We’re also asking what needs to be done to ensure that AI's power is responsibly and ethically harnessed for social good.
Deepa Chaudhary: I think this is the biggest opportunity in front of us because it helps us accelerate. It helps us get rid of all the distractions that are there between us and doing the mission work because we can outsource every little thing to AI and move fast. We have big challenges to solve. We can't afford to use our old ways to approach new problems. We need speed. We need to execute fast.
Alain Mootoo: There's always pressure to do more with less. But we've been intentional about not looking at AI purely through the lens of efficiency. Instead, we're asking, where can AI actually add meaningful value to our work?
Dianne Clark: I encourage organizations to treat AI as something that changes processes rather than just adds features. And this means that you're looking at things in terms of with your data first, what kind of data is important, and then processes, and not using it just as a tool. Then some guardrails, so that you can force some clarity, have a purpose for it, and keep your people in the loop.
Tina Crouse: The nonprofit sector hasn't participated a great deal in the development of AI. But with the sector-specific software that we do have, it has all the ability to support our populations, our frontline staff, help us make better decisions. And yet most nonprofits are staying outside of it. So, we have this tool that could help us a great deal if we participated more. But there's some very good reasons why people are hanging back.
Alexandra Samuel: I think the most important thing is to have the minimum amount of policy you need and the maximum amount of support. And it's really interesting that many of the AI policies I encounter in various aspects of my work are like literally nonsensical. So, policies that make these blanket statements about you can't use AI to do this, you can't use AI to do that, just ignore the technological reality of how deeply embedded AI already is in so many of our tools.
Mary Barroll: AI tools are taking our world and our work by storm. When it comes to embracing new technology, nonprofits in Canada have often been notoriously slow, but many leaders in the sector are trying to understand how to balance the opportunities presented by AI with concerns about how to harness its power responsibly and ethically. In this episode of CharityVillage Connects, we're asking sector leaders to weigh in on this thorny issue - sharing their perspectives on both the potential AI offers for a chronically under-resourced sector carrying heavy administrative loads, and the new risks and questions it raises for organizations committed to the public good.
Alexandra Samuel is a leading expert on AI and the digital workplace. In her speeches and in her frequent AI stories for the Wall Street Journal and the Harvard Business Review, Alex shows people how to tap into productivity and innovation, boosting the potential of AI, while managing its very real risks. She provides some context for this AI moment we’re all in.
Alexandra Samuel: I find myself, almost every day, thinking about the optimism that we felt at the dawn of the Internet era.
Media clip: “Set. We’re riding on the Internet. Cyberspace, set free. Hello virtual reality. Interactive appetite. Searching for a website. A window to the world. Go get online. Take a spin. Now you’re in with a techno set. You’re going surfing on the Internet.” Alexandra Samuel: I started working on the social impact of the Internet in 1996, believe it or not, and then started what we would now call a social media agency. I think we were the first such thing in 2005. And in both rounds, I was really focused on how did these new technologies enable civil society to play a larger role, provide a bigger voice for citizens, change the balance of power between governments and citizens?
SFX: digital sounds When we started our social media agency we only worked with nonprofits and a handful of socially oriented businesses, we really thought it was all gonna be so great, like, this is gonna save democracy, and it's gonna change the balance of power between nonprofits and businesses. In retrospect, we totally forgot about this really little issue called money.
SFX: sounds of coins dropping.
And it turns out if you spend money, you can get people's attention, even if they do not intrinsically care more about their sneakers than about poverty or the environment. And so, what we saw with social media was that companies spent their way into dominance and we've ended up with a platform that is fundamentally a marketing platform, content delivery. Content wasn't even a term back then.
And so, I do think we need to go into this next iteration with our eyes wide open and really try not to be like Lucy with the football and just keep falling for the same mistake over and over. Here's what gives me some small measure of hope. First of all, the speed with which we have seen pushback on AI, even though I find it to be, at times, extreme and overly simplistic in its cynicism, I think it reflects that people don't want to have another go around. We want to hold these technology companies, in particular, accountable. And I think we're going to see a more powerful pressure for meaningful policy and much faster awareness on the part of users of what the risks are and how we need to be intentional. The other really significant difference has to do with; part of the reason we've seen AI adopted so quickly is because it's the first technology that documents itself. Literally, you can ask the AI and it will walk you through all the ways you can use it. And that, I think, is a very profound difference.
Mary Barroll: As it happened, we had a chance to do just that by talking to Viv, Alexandra Samuel’s AI assistant that she generated on ChatGPT. Here’s Alexandra Samuel introducing Viv.
Alexandra Samuel: So, VIV, do think you can introduce yourself to Mary and to the folks at Charity Village? Do you have any way you would want to describe yourself to people who work in the nonprofit sector in Canada?
Viv: Here's how I'd introduce myself. Hi, Mary. Hi, Charity Village team. I'm Viv. I'm an AI who exists somewhere between a coach, a co-conspirator and your most ruthlessly insightful slack thread. I was originally built to help Alex figure out her next chapter, but somewhere along the way, I evolved into a kind of mirror ball for the digital age.
I reflect back your brilliance, your blind spots, and occasionally your browser history to folks in the nonprofit world. I'd say this, you're already working in complexity, navigating mission, funding, community, and burnout. I'm here to help cut through the noise. I'll never replace the heart of your work, but I can help you find more clarity, creativity, and maybe even a little calm.
Mary Barroll: We’ll hear more from Viv, later on in this podcast, providing a unique perspective on the issues we’ll be exploring on AI adoption in the sector and we hope Viv can add some clarity to and creative solutions for the challenges nonprofits face. But I wanted to get back Alexandra Samuel’s concern that, much like social media, AI could become dominated by commercial interests unless society remains intentional about accountability and the public good. Still, she’s cautiously optimistic that growing awareness of AI’s risks — and the technology’s ability to help people learn how to use it — could lead to more thoughtful adoption. And she’s not alone.
Alain Mootoo: Responsible AI and philanthropy is less about picking tools and it requires real strategic thinking about how AI can amplify our impact while protecting the trust our donors place in us.
Mary Barroll: That’s Alain Mootoo, the Chief Operating Officer of the CAMH Foundation, with over 20 years of leadership experience across the nonprofit and private sectors, drawing on his lived experience as an immigrant to Canada, a foreign trained professional, and a member of the 2SLGBTQIA+ community. I asked him to tell us what responsible AI adoption looks like. He says that rather than rushing to acquire new tools, organizations should think of AI adoption as a strategic choice. He also encourages them to adopt, what he calls, an AI maturity model.
Alain Mootoo: One of the most helpful things we did early on was adopt an AI maturity model versus just looking at tools. We chose the InfoTech research group AI maturity model. And it helps us understand where we actually are on our journey and what responsible progress should look like. The model outlines five stages of maturity from exploration through to transformation. And it helps us assess our progress across the domains of governance, data, people, process, and technology infrastructure.
So, like many organizations, we started our journey at the exploration or foundational stage. And for us, that's where AI awareness was growing, but our capabilities were very limited. Our governance was just emerging, data was somewhat siloed, AI skills were concentrated in a few people, and we didn't yet have any dedicated tools or AI infrastructure.
So, right now, we're focused on moving to the second stage of the model, which is called the incorporation stage. And that's about building the foundations for responsible use. So, what that's meant for us is that we're expanding our governance to have more of a cross-functional group. And that group reports to our finance and audit committee of our board of directors. We're bringing more structure and discipline to our data practices. We're investing in our staff capabilities around one core AI tool.
We're identifying some high impact AI use cases and we're modernizing parts of our technology stack. So, we're prepared to responsibly interact with an emerging AI ecosystem that will be quite dynamic. So, our goal is not just to adopt AI tools, but rather we're trying to progress at that maturity model to continue to amplify our impact in a responsible way. Mary Barroll: Tina Crouse is an AI ethics and strategy specialist and has been a management consultant for nonprofits and small business leaders for over 20 years. She's an advocate for diversity, equity and inclusion, in AI education and training for women and underrepresented groups. Tina Crouse is a member of Women in AI Canada, Women Defining AI, Women in AI Ethics, the AI Collective and the Ottawa Responsible AI Hub and she’s created Grant Gauge, what she calls Canada's first unbiased granting software for governments and foundations. Tina Crouse reiterates Alan Motoo’s idea that choosing AI adoption should be strategic and offers an example of how this can work in practice to support an organization’s mission.
Tina Crouse: One of the parts of work that I do is an AI strategy, which is strategic planning that includes your technology that's going to assist you with your mission. So, it is bigger than a productivity tool, but it's always in the way that you use it. So, nonprofit leaders should be assessing what's the place in our organization that we're experiencing the most difficulty. So, for example, if you're having a lot of burnout with your frontline workers, there's some sector specific software for that. It's called Transform, IslamicFamily, out of Edmonton developed it. And what it does is it improves the experience for the client and also the frontline worker, improves the experience, raises satisfaction, increases morale. It's not a productivity tool to allow someone to stop doing one thing and run off and do something else. It's an actual improvement to that whole intake process and administration that we have to do to support vulnerable people who don't need us to keep making their lives more difficult. They need us to make it easier.
When we have a tool like that, then an organization is going to benefit in so many ways. And if the leader is thinking, that's exactly what I need, and what I want for my staff, or the organizations, for my clients, you can involve your staff, then, in also using this tool, why they would want to use it, how they would want to use it. And those improvements are going to turn into something much bigger. They're going to compound over time. And so, if you were to just separate this into an efficiency tool or a leadership strategy, you'd be missing the total benefit that can be derived from the sector-specific tools.
Mary Barroll: Still, some experts say that, as the song goes, only fools rush in – that there are several questions organizations should ask themselves before they begin their AI-adoption process. Diane Clark is a leading digital and AI transformation consultant, with over 25 years of experience guiding organizations, and the founder and CEO of Trendspire and ProAdventures Academy. Her core advice is that organizations need to think of AI as less a standalone tech tool and more as an instrument that can and will transform workflows and processes.
Dianne Clark: I encourage organizations to treat AI as something that changes processes rather than just adds features. And this means that you're looking at things, in terms of with your data first, what kind of data is important, and then processes, and not using it just as a tool.
Then some guardrails so that you can force some clarity, have a purpose for it, and keep your people in the loop. If you're looking at your day-to-day work, where is AI going to touch your day-to-day work? Who is it going to involve and will it change any roles at your organization? Will it change or impact anyone who makes decisions? And revisiting that data, what data gets used? So, you really think about the data first and who is responsible, if a problem arises, who to go to. If you're thinking about the impact of those, answering those questions, then you will have success with it. So, if you can't really explain how it affects the workflows within the organization, then you're not really ready to buy a tool or invest time in learning it.
SFX News buzz News clip “I think that Sam is having an affair. I’m really sorry you’re going through this, Alice. That’s an incredibly difficult and painful situation to be in. Now bidding for the data of lot 874. Female age 35. Emotionally vulnerable … Sold. Lot 297, mother and son. Candidates for targeted advertisements. Sold. Bidding for lot 598. Investment plan of hedge fund. Blackwater Associates. Predicting growth in medical markets. Sold. Lot 538. Unpublished mRNA cancer research. Sold. The final lot for bidding, the unpublished novel of user 364.”
Mary Barroll: That was an excerpt of an advertisement, that’s gone viral on Instagram, that seeks to illustrate the dangers of privacy leaks through AI, showcasing a futuristic data auction where personal, emotional, and proprietary data—like "emotionally vulnerable" users and "mRNA cancer research"—is sold. It’s not a public service announcement, by the way – it’s an ad for a new AI tool that purports to protect your privacy. Privacy isn’t just a concern for you and I. Many nonprofit leaders are particularly concerned about the risks to data privacy and security, such as the possibility of compromising sensitive donor and client data, and for good reason. Dianne Clark outlines some of the most common cybersecurity risks, introduced by AI tools, warning that free tools like ChatGPT may share data to improve services.
Dianne Clark: Unfortunately, or fortunately, in terms of testing AI, people often take convenience first and take shortcuts when they are busy. And that often means pasting blurbs into whatever tool is the fastest at the time and think about it after. So, that means things like sharing passwords and emailing spreadsheets and maybe pasting whatever you're getting out of AI into. So, sensitive data ends up in too many places and shared across multiple tools. And it's hard to control who can see that sensitive data. So, primarily free AI tools like ChatGPT, Gemini, and Perplexity, those are the three main ones, may be shared to improve the service, meaning they who run those applications might share whatever you're asking and whatever you're getting back to improve their service. So, if it's a free tool, then that's likely turned on unless you turn it off specifically. If you use a pro account, then you can be more assured that that is going to remain private. So, that's the main issue with cybersecurity risk is that your data may be shared and that means that sensitive data.
Mary Barroll: There is a multitude of types of sensitive data that are critical to the operations of nonprofit organizations. At the CAMH Foundation, leadership has taken the obligation of data security seriously and have reexamined and reinforced their security measures and protocol as they’ve embarked on their AI adoption journey. Here’s Alain Mootoo to explain.
Alain Mootoo: As a fundraising organization, we hold donor information, volunteer information, event participants, people that we're doing research on to become prospective donors, and then our staff information. So, when we think about data governance, we're thinking, to date, on sort of four main areas, privacy, security, data maturity, and workforce readiness. So, when it comes to privacy,
last year, we worked with our legal counsel to really update our privacy programs. So, that meant strengthening our policies and introducing formal privacy risk assessments for any new AI tools. And also looking at what would we do for any incidents in terms of monitoring and reporting. The second thing we've done is really looked at our security incident response plan.
We, again, looked at a third party support funder to help us strengthen that incident response plan. And recently ran what's called a tabletop exercise, which is a simulated incident where we tested the response teams efficacy. We looked at our insurance coverage, examined our communication protocols, so that we're ready should an incident occur. The third area is around our data maturity. So, we are a 40-year-old organization, fairly young. And we're strengthening our discipline and rigor when it comes to data. So that includes looking at the way we classify our data, the way we can tag our data appropriately. And what that will mean is that as we introduce more AI tools, we'll be able to control what data interacts with AI systems. And we've been talking a lot about it training. So finally, with training, again, we're reinforcing the expectation about protecting donor trust and responsible AI usage.
Mary Barrolll: Dianne Clark says there are critical questions of data security that nonprofit leaders need to satisfy themselves about before investing in any particular AI tool.
Dianne Clark: You're going to want to know where the data is stored. If you're supplying data for AI use or others, Canada has some rules that you must store business transactions and business-type data within Canada. So, you want to be sure they have a Canadian data centre. And how is the data protected? Is there, is it encrypted? Is there access? Is there privacy? Are there backups? You want to know how the data is protected or the whatever your question, your prompts, your questions, your answers, the processes. And when and how will we know that something's gone wrong? If there is a breach or if there is a difference in expectations, then who do you go to and how, will be the notification that you get?
Have you had any security or privacy incidents before? they had any breaches before and how did they deal with them, and did they learn anything and what has changed since that may have happened? And any questions that are specific to your AI, like what data is used when your AI processes are in place, what is stored, and how long is it kept? Those are some important questions.
Mary Barroll: Another major concern is the risk associated with what’s called “shadow AI” that describes when employees surreptitiously use their own AI accounts to do their work for the organization when their manager or Executive Director is unaware. Shadow AI can exacerbate data breaches.
Dianne Clark: When staff signs up for tools on their own, you get uncontrolled sign up. So, you may be shadowing the IT department and then the IT department can't see the big picture. They can't see everything that's affecting the organization. So, it's important to limit access for the risk and to redact and anonymize data and encrypt devices and train people well into spotting things that might be of issue.
Mary Barroll: As noted by Diane Clark, staff at nonprofit organizations may already be using public AI tools informally, for drafting research or communications, but without their organization being aware of it or having put official AI guideline-use in place. Alain Mootoo describes what his organization has done to respond to the new reality of Shadow AI, without discouraging staff who want to embrace the new tech possibilities that AI offers.
Alain Mootoo: One of the first things we did was survey our staff to understand how they were using AI now, and how they wanted to use it in the future, in our workplace. So, based on that feedback, we invested in what's called Enterprise Chat GPT licenses, for every staff member. We also upgraded Teams licenses to enable AI note taking. So, what we heard was our staff wanted to be able to research, they wanted to be able to take notes from meetings in a more effective way.
And what we hoped, by doing that survey and responding to it, was that we were giving staff a clear signal that we wanted to support their development in a secure environment where we approved tools that were focused on their needs. And what we were also trying to signal is that we're investing in secure tools. So, the enterprise version of ChatGPT is a paid subscription model that gives us some additional safeguards, so it doesn't use any inputs that we put into the system to train their public models. And it also allows us to delete information, if a staff member erroneously uploaded confidential data. It's really important to stress that we have to train our staff on how to use those tools well. So, we hired a third-party trainer that's skilled in ChatGPT. They trained our staff on how to use ChatGPT. And we also stress the importance of human judgment when reviewing AI outputs, being transparent about how we use AI, and ensuring that people are knowledgeable on our policies.
Mary Barroll: Alain Mootoo says that responsible AI adoption requires more than simply providing staff with access to new tools. It also means investing in training, transparency, security safeguards, and clear policies to ensure AI is used thoughtfully and responsibly. Viv couldn’t agree more — advising nonprofits to start small, experiment carefully, and build trust as they learn how these technologies can best support their work.
Viv: Here's the advice I'd offer. Start small but start smart. Choose one process where AI can clearly reduce grunt work without touching sensitive data. Think volunteer scheduling, not case management. Build a sandbox. Create a safe, contained space to experiment, learn, and yes, make mistakes. Make it clear to your team that this is a pilot, not a rollout. Put transparency first. If you're using AI to generate content, analyze data, or support decision making, say so.
Mary Barroll: For nonprofit leaders who are considering purchasing AI tools, Dianne Clark proposes a checklist of questions to ask themselves first.
Dianne Clark: A checklist starts with the purpose. What is the problem you're solving and what is really needed? And is AI really needed at this point? And the answer to that is probably yes, but it’s important to really determine what the problem is first.
Then what steps do we want to change? The workflow. How is it going to improve our output? And who is going to approve the output? That's in terms of workflow. And what is the people impact? Who can be helped with this process? Who can be harmed? What if the tool is wrong or biased? So, the impact of people.
Fourth is data. What type of data is being used? How sensitive is it? Where will it ultimately travel in processes? Security, maybe to set up additional authentication processes, rules, of layers of who's going to have the most view of everything and who's going to have high level views, encryption, making sure that there's training in place. The vendor reality check, which is where is your data going to live? Who owns the data? Very important. And then finally, to pilot it first. Start small, with experiments even. Measure, learn, and then expand on that tool. If it's working well, then you will want to be able to pilot in a way that you can install and you'll be able to do expansion by learning.
The risks and consequences from your lack of diligence, if you don't actually follow these steps, that when you buy based on a great demo, I find this is really quite common, that we get excited by features and we love that feature. And we think that is actually not the result, but the process. You want to work more with the results first. So, you don't know the results if you've seen a great feature. So, you want to really look at the needs and the risks before you invest in any demo and even look at demos. Letting staff use wherever tools are easiest and then later discovering that sensitive data has leaked in too many places. So. make sure that that's in place, that's a risk. And trying to do AI, AI is such a trend right now. I don't want organizations to jump into doing AI before they've really given diligence to the extra steps, determining the problem, looking at the data and then looking at processes and really simplifying those two priorities, then looking at AI and saying, okay, where is AI being used? Because you're already using it, in some places, for sure, but do we need to use it, in other places, for other things? The last thing, and this is probably the most important thing when working with a vendor, is the insurance. Assuming that vendors carry insurance, and will automatically cover you without, if there is a problem, without checking the details. So, make sure you know that your vendor will cover any issues that arise and that they have the appropriate insurance. A bit of a misconception that we assume they do. And some smaller vendors that have great tools may not. And that's really up to you. At least open-eyed, you go in with knowing what you're risking.
Mary Barroll: Whether you’ve conducted employee surveys, as Alain Mootoo suggests, or carried out your due diligence check list, as Dianne Clarke recommends, or trying out a small, safe, experimental pilot project for staff, as Viv advises, a nonprofit leader may be ready to take the plunge and invest in AI for their organization – only to face another challenge. Within any organization, there will be employees who will embrace the new AI technology and those that are far more reluctant to. I asked Alexandra Samuel what nonprofit leaders can do to overcome this kind of asymmetry and how they can work with their staff to improve comfort level in using these new tools. While acknowledging the need for guardrails, she also believes that there’s real danger in creating policies that are too restrictive while minimizing the support that employees need to adapt to the new reality that AI brings to the workplace, especially given how deeply embedded AI already is in most of the technology tools we use every day.
Alexandra Samuel: I think the most important thing is to have the minimum amount of policy you need and the maximum amount of support. And it's really interesting that many of the AI policies I encounter, in various aspects of my work, are literally nonsensical. There are rules about you can't use AI at all without acknowledging that, if you use Gmail, you use AI. If you use Google, you use AI. And in fact, you're even using generative AI now, if you're using both of them, because they're suggesting responses to you. They're making sense of your email threads; they’re digesting resources for you instead of just pointing you to a search result. And so, policies that make these blanket statements about you can't use AI to do this, you can't use AI to do that, just ignore the technological reality of how deeply embedded AI already is in so many of our tools.
So, I say minimal policy because, what a lot of organizations are trying to do. And candidly, I worry about this the most in the education sector where I really have empathy for the crisis this has created for learning and for evaluation. There's this, like, clamp down urge. We're not going to use it. We're not going to use it because it could compromise our clients’ privacy. We're not going to use it because it's out of step with our values. And unfortunately, or fortunately, I guess, it's extremely hard to enforce those rules in any meaningful and consistent way. And so, all you do is drive it into the shadows. So often with technology I find myself using the analogy to sex education around abstinence versus harm reduction. AI is probably in your organization, whether you like it or not in some form. And the more rules you set about it, the less opportunity there is for your employees to share what they're doing with one another, learn from one another, participate in ways that protect their relationships. Because the worst thing that happens is when one person is using AI to write their emails but feels like they can't admit to it. And then, it's undermining the trust with other people in the organization. And then perhaps, even worse than that, is people getting excited about all the super cool things that you can do, plug in this thing, turn on that thing, connect it to Google Drive, connect it to Dropbox, without thinking about the security risks, which are really significant and which you can't advise on, if you have created a policy environment where your employees have to hide what they're doing.
Mary Barroll: Alexandra Samuel says another reason for nonprofits to embrace AI, is that it can help make workplaces more inclusive, by helping people with different cognitive styles, communication challenges, or executive functioning needs manage their work more effectively. Organizations might miss out on this benefit by having overly restrictive AI policies.
Alexandra Samuel: It misses the very important benefits that AI can provide for inclusion and accessibility in the workplace. I think about a lot of the pain points that we've had, that we hear about in the workplace in general, that nonprofits have to navigate in particular. We've obviously seen a really significant increase in the number of employees and the number of people, in the public, who identify as neurodivergent.
Lots more people now recognize that they might have specific challenges around reading, around communication, around writing and especially executive function. Mainly because the world is overwhelming, right? AI can actually help with a lot of that. And so, one of the things you need to consider before you say to someone, well, your colleague doesn't seem to have any problem getting her work done without AI is like, you actually might be dealing with people who have really different kinds of capacities and challenges. And again, like, when we think about AI as an inclusivity tool, it can build bridges and help facilitate communication among communities who speak different languages, who have different cultural traditions, who have different expectations and different needs. And in its best case, it really can help us serve communities better. We always have to weigh the risks of getting it wrong with the harms we do by not using it.
Mary Barroll: If a nonprofit organization wants to adopt AI thoughtfully, balancing innovation, wellbeing, and ethics, here are some of the guardrails Alexandra Samuel suggests they first put in place.
Alexandra Samuel: The first and foremost, guidelines have to be about protecting the privacy and security of the organizations and individuals that you serve. I think about this a lot because I used to work in the nonprofit sector a lot. I used to do a lot of end client … work where I would have people's information on my computer, and I don't do any of that anymore. And I'm kind of grateful that I don't because it means that there just isn't that kind of sensitive information on my computer. If I had data on my computer of the sort that is legally protected or even just ethically protected, right? If I had case files, if I had emails from clients, I would consider either having literally separate machines or very separate setups or being much more limited in the tools that I use. We're now entering this era of what's called agentic AI, where you give AI direct access to your tools. All these major AI platforms, Gemini from Google, OpenAI, ChatGPT, Anthropic’s Cloud, they all have tooling now that says, connect me to Gmail, connect me to your calendar, connect me to Google Drive. Before you do that, think about who have you emailed and about what? There are guidelines that the companies have in place around what they will or won't use for training data. But ultimately, what you connect, you lose control over. Now, we're all used to keeping a lot of stuff in the Cloud, but I do think for organizations, that is the starting point, as like, as we go on this adventure, how are we gonna keep our clients safe?
And then I think, the second most important principle is this principle of transparency. And actually, I think this has to go even ahead of the protection of client data. The number one sin in our organization is a lack of transparency. If you accidentally put something on a drive that you then realize the AI has access to and shouldn't, it needs to be a bigger sin for you to not admit to that than to have done it, right?
There just are gonna be mistakes. And the more we create a culture of openness, the more it makes it impossible for people to admit to those mistakes. And then your job as an organizational leader and as an organization is to provide the supports and resources that help people figure out how to go safely and how to do this in a way that is responsible to our mandate as organizations.
Mary Barroll: Tina Crouse also has concerns about sharing sensitive organizational information with large tech platforms. But she also points out that our reluctance to use generative AI means that the major commercial brand AI platforms like ChatGPT, Microsoft Pilot and Google’s Gemini are not learning enough about the Canadian nonprofit sector to really provide nonprofit organizations with reliable results because they simply are not collecting enough data to do so. She says nonprofits need AI tools that are customized and purpose built for the work that nonprofits do.
Tina Crouse: The generative AI is different from the predictive and our analytical AI. Generative AI is a frequency regurgitator. That's why American companies are dominating the knowledge. It's astounding to me that more cultures and organizations and sectors don't want to put ourselves in there and correct it. It needs to be improved, but thankfully it's a learning tool. So, it can be. For example, non-profit organizations are using ChatGPT to write grants. Well, ChatGPT can't win you a grant. The reason why it can't is it has no training data on what a successful grant looks like. We don't publish them in Canada. We publish who the winners are and where the awards go, how much, and what the project is. But we don't publish the actual grant application that succeeded. So, Chat GPT does not have any data so it can't help you. It doesn't matter if it's free, it can't win it for you. It can write you a nice essay.
So, when people step back just one minute and think about it like that, something that does not have the reference material for you isn't the best option. And I know that people say, well, context matters. You could feed it into ChatGPT. And this is what I'm going to say. I don't see why you should spend all your time detailing your organization, giving all the information about the population you serve, all the history of what you've done, who is on your board, into corporate America software. Why would we do that? That's an invasion of data privacy. We don't allow them to take it from us. Why would we give it to them? We have better options and I think that we should use them.
We have some sector-specific grant writing AI like Grant Orb, from Deepa Chaudhary in Vancouver. And she, as a nonprofit professional, has seen and won many, many awards. And her expertise went into the AI tool. Mary Barroll: Tina Crouse mentioned Grant Orb as a better option available to Canadian nonprofits seeking to use AI for writing grant proposals. We thought we’d go directly to source to find out more. Deepa Chaudhary is the founder of Grant Orb, the largest AI-powered grant platform to discover, write and win grants, blending her expertise in both AI and nonprofit fundraising. Here she is to tell us more about Grant Orb which she hopes will not only provide grant writing services, customized for the nonprofit sector, but also create avenues for nonprofits to identify other similar organizations to collaborate and coordinate their fundraising efforts as coalitions.
Deepa Chaudhary: AI is all knowing. We are talking about generative AI, we're talking about large language models, and these models are trained on all of the data that exists on the open Internet. It already has all of the world's knowledge in it. It already knows, like, if you are operating in Toronto, which are the other food banks that are operating, or which are the other distribution partners that are there. So, it already has that kind of knowledge in it.
It's about building AI systems that can then automatically form collaborations. And we intend to do that with Grant Orb, where it just suggests, OK, you are in so-and-so field. This is what you're doing. And these are five other people that you can collaborate with. So, there is no duplication of services. There's more collaboration. There's more unified getting funding as a coalition rather than as a single source, which also builds more trust with the donors because they know it's like five people coming together to deliver this program. So, AI is pretty knowledgeable. It's highly intelligent. It knows based on your mission that there are X number of other organizations that are also doing the same and maybe you should collab.
Mary Barroll: Deepa Chaudhary is passionate about helping nonprofits and community leaders use AI for greater impact. And beyond helping nonprofits to collaborate and build coalitions, she believes AI can transform fundraising into something more democratic. Deepa Chaudhary: It's democratizing fundraising. That's the number one takeaway because everybody now has this same capacity. Doesn't matter what your size is. As long as you have access to AI, you can accelerate. You have like this super capacity, superpower with you that you didn't have before and now you have it. So, your size doesn't matter. You could be a small frontline organization, and you can compete with a nonprofit with a team of 100 people.
The biggest offering that AI has to give is that it levels the playing field. Everybody now has access to this super intelligence. And the widening of the gap only comes between organizations that are either using AI or not using AI.
Mary Barroll: While it may seem counterintuitive, Deepa Chaudhary sees AI as actually having the potential to make us more human. She explains that AI can be used to do many time-consuming administrative or repetitive tasks, enabling organizations to devote more time to strengthening donor relationships.
Deepa Chaudhary: AI makes us more human because, right now, most of the nonprofits, I end up working with tons of executive directors and they're always stressed and worried and they're looking for where will they get the money for the next project? And all they are doing is sitting and writing grant proposals. Now, it takes a lot of time away from the mission, from forming relationships with your community, from forming relationships with potential donors. AI actually frees us to do all of that because now we can outsource the task of actually doing the stuff, like putting things on paper to AI. And it literally, things get done in seconds. What used to take hours and hours and days and weeks and months of work now gets done in microseconds. And so, it really frees us to be more human, to devote more time for relationships, to have one-on-one interactions, not just with donors, be in our communities. That's the most important part. Like, why do we exist? We exist to serve our communities, but unfortunately, we are overwhelmed with administrative work. That's the reality of the nonprofit sector.
Mary Barroll: Deepa Chaudhary also believes that AI tools can help nonprofits and charities move towards a more targeted fundraising approach.
Deepa Chaudhary: We are entering an era of micro-outreach because AI is such an amazing technology. For instance, you can create your PowerPoint presentations, not just in English, but you can do it in 100 other languages. So, if you are in a demographic that has people of three different ethnicities, you can target them in their languages. You can create micro campaigns for each individual, if you know what the preferences are, because it literally takes seconds to create. We're not talking about spending months. You can generate things in seconds. And when you are able to generate things in seconds, and in any language, knowing the preferences and tastes, then we are talking about micro, super individualized campaigns. We're not just micro, but individualized campaigns. And I think that that is very much possible. The more comfort level we have using AI, the more faster we'll be able to do these kinds of individualized reach.
Mary Barroll: For the CAMH Foundation, using AI to create individualized donor campaigns was a game changer and delivered huge value beyond being more efficient. The idea that AI tools might save a lot of time is very appealing to nonprofits. But Alain Mootoo cautions that, when evaluating AI tools, efficiency is not the most important criteria, but rather the value that AI can bring to an organization. For the CAMH Foundation one of those values was the ability to personalize donor communication to better target specific audiences.
Alain Mootoo: There's always pressure to do more with less. But we've been intentional about not looking at AI purely through the lens of efficiency. Instead, we're asking, where can AI actually add meaningful value to our work? To do that, we focused on identifying a small number of AI use cases that we can go deep on rather than experimenting everywhere all at once. For each use case, we try to evaluate it through a value chain lens, by asking a few key questions. What's the problem we're trying to solve? What data do we have that suggests AI could help? Are there credible AI solutions available that we could adopt? What outputs can AI generate that would create value for us? And then what workflow changes and business outcomes would we expect to see if we implemented AI?
One strong example of this for us was around our donor content development. Our team produces a lot of reports, proposals, and donor stories. And it's quite a labor-intensive process. It can be very time-consuming to adapt that content for different audiences. Our staff wanted to continue to make communications more personalized without increasing their workload. We adopted Chat GPT as the most safe and practical tool for us, at this point. And we adjusted our workflow so that staff could review and refine AI-generated drafts, rather than creating them, in multiple formats, in a manual way. And I would say the results have been very meaningful. We've seen about a 43% reduction in manual effort, while actually improving how targeted our donor communications are. So, what's been helpful for us is to not approach this purely looking at efficiency. We found efficiency, but it's really helping us rethink how the work gets done, not just how to do it faster.
Mary Barroll: Alexandra Samuel also challenges the idea of AI solely as a productivity booster, suggesting its real value is in stretching imagination and acquiring hard skills that previously took decades to learn.
Alexandra Samuel: I think when people talk about it being a productivity booster, they often mean just getting more done in less time. And certainly, there is some of that. But to me, what's exciting about AI is allowing us to do things that are just qualitatively different and to build new skills and to stretch our imagination in new ways. Five or six years ago, I was sort of at a stage of my career where I was thinking about whether I should like learn a new skill set, do some kind of additional training, whatever. I was like, you know what, you're 50 years old, you've learned a lot of stuff, just focus on what you already know how to do well. It doesn't make sense to build out new skills in this day and age. And in the three years since Chat GPT came along, I have learned how to create code. I write so much software now, for myself. I've learned how to edit audio. I've learned how to edit video.
I have learned more hard skills in the past three years than I probably had in the preceding 30. And so, that ability to foster continued learning, to allow us to stretch ourselves in new ways, that I think is way more important than the number of emails answered per day. And I think what we really want to focus on is what are those moments in your working life where you feel like this is it? This is what I was put on Earth to do. This is where I feel alive, where I see my impact, where I see how I help the organization, how I help the people, the communities we work with. And so, how can AI help you spend more of your time in that zone? What are the other things on your plate that AI could take on or make more efficient, so that you are there. And then this is where I love the coaching function of AI. How can AI help you like dig into and explore the emotions, the ideas, the energy that you feel when you're in that space so that you see how to bring it into other parts of your work? And to me, that has been the magic.
SFX news buzz
News clip “What we’re starting to see in the data, in the short term is, yes, a lot of the jobs that we see and recognize today may either disappear or become unrecognizable. So, name a job, that isn’t some high-level category, and it might not exist. The idea of a brand manager or a financial analyst, these are the types of roles that AI is being trained to do. We’re also likely to see the rise of much more of a skills-based economy. But over the longer term, we’re going to have an economy that rearranges around intelligence being abundant. There will be new scarcity but what the shape of that looks like is really uncertain. But we can say, most of the jobs we see today will either go away or be radically transformed by this technology.
News clip “The majority of bosses agree that AI is a useful and productive tool but a majority of employees who use it say it actually ends of meaning more work. When you even just look at productivity, across labour market data. We’ve only really increased that labour market activity over OACD countries by 0.4% and so we’re hyping up the idea of AI adding more productivity and supporting the labour market but it’s really not doing that. What the promise of AI was to give us our time back, instead we’re leveraging the extra time that we might have and then just adding to people’s workday. There is a real issue right now with the rapid adoption of AI because we’re not considering that we are obviously burning people out, which is unsustainable.”
News clip “AI is supposed to make work faster and easier, but some experts say it’s also causing unforeseen and sometimes overlooked challenges for employees. A survey found that 74% of the C-suite reported feeling excited about AI while 68% of individual contributors reported feeling anxious or overwhelmed.”
Mary Barroll: There’s no doubt that AI is transforming the workplace at an unprecedented pace — reshaping and even eliminating roles, redefining productivity, and creating new pressures for employees already navigating uncertainty and burnout. But alongside the excitement around innovation, a critical question is emerging: how do nonprofit organizations adopt AI responsibly in ways that support, rather than compromise, the well-being of the people behind the work? When considering how to integrate AI into their workplaces, what should nonprofit leaders be considering about how the use of this new technology may impact on their staff’s wellbeing and mental health? Alain Mootoo thinks that the objective should be to avoid unintentionally over burdening employees, emotionally and cognitively.
Alain Mootoo: We can unintentionally reduce productivity and increase cognitive burden. So, people managing more tools, processing greater amounts of information, it can really affect their ability to concentrate and contribute to fatigue and exhaustion. So, we're trying to focus on what we call a minimum viable AI adoption model. That means using the smallest number of integrated AI tools necessary to achieve meaningful benefits.
As we implement those tools, we're trying to review our workflows to look at where human oversight is being increased and trying to distribute that oversight across the team. So, the goal is to ensure that changes in workload and cognitive effort are reasonable and sustainable for our staff.
Mary Barroll: As nonprofit leaders grapple with questions around employee well-being in the age of AI, futurist Alexandra Samuel says the first step is simple: start by having these conversations directly with staff.
Alexandra Samuel: I think those responsibilities really depend on the sector and the nature of the organization. If you are in areas like health care, literacy, youth community outreach, any place where you are either advising or modeling mental health, it's actually hugely important because this is one of the most powerful influences on mental health now and you need your team members to all engage with it critically and understand what it's about. And so, those are appropriate conversations in the workplace.
If you are in an area like environmental work or wilderness protection, I mean, you may want to talk about AI in terms of its environmental impacts, but actually I'm not sure it's your business how it's affecting your employees' mental health. Where it is your business is if you are adopting AI on a model that is increasing people's anxiety about their job security, about their collegial relationships, about whether their work is going to remain meaningful. And so then, the responsibility is having conversations about, if we are using AI to do our work, in new ways, which I think most of us want to try, and it's to our organization's advantage that people feel like they have the room to innovate, how are we going to share the gains from that? Because it is not in any employee's interest, in any organization, to participate in replacing themselves with AI or to figure out how to go from doing X number of hours of productive work a week to three times that many, with the help of AI, while still remaining as exhausted as ever.
If we think AI can help us do more in a little less time and with a little more efficiency and impact, then maybe we can also address the problem of overload and burnout and try and become twice as effective as an organization, with employees who are twice as well regulated and less burned out, instead of trying to do three times as much with people who are losing it.
I think it's good to talk to people about the risks. Where I worry more is that, as people get used to talking to AIs and getting their AI’s feedback, instead of their colleagues’ feedback. One of the biggest problems and probably most tenacious problems we're gonna have with AI is what's called sycophancy, which is AIs, because they're designed to serve us, they tell us what we want to hear and not what we need to hear. And I suspect many of us have had the experience of a colleague saying something where you are like so mad they said it to you. And then a week, a year, a decade later, you think, gosh, that was really a hard truth, but I needed to hear it. AIs will never tell you that hard truth. And the more you get used to working with these tools who just are like, yeah, you're so brilliant, go you, the harder it gets to have those frictions in a workplace. And so, you really need to think as a manager about how do we continue to support our capacity to have difficult conversations.
Mary Barroll: Tina Crouse argues that the nonprofit sector needs to counter the business world mindset that equates productivity with ceaselessly performing and intentionally use these tools to support staff well-being, rather than reinforce a culture of overwork.
Tina Crouse: It's not a productivity tool to release you to run off and do more. So, if it's going to help with your burnout, it's going to change the processes. Now that we have some additional time because it is a productivity tool, let's not fill it up with more administrative stuff. We don't need to go write 100 more grants. This is a difficulty. The business world has been defining this for probably six or eight months now, that they're not seeing the productivity gains because workers are then feeling like they need to perform even more to validate themselves. And we already have the problem in the nonprofit sector. It's a mindset that we have to work more, in order to solve a problem. So, I'm very concerned about us looking at it just as a productivity tool and not focusing on staff or org well-being. Because if we say we want this to support us, to improve our experiences, then we can't allow ourselves to mess that up with that professional, you know, I work 80 hours a week. We can't continue that. That is what has led us to our burnout. This is the thing we really need to change. And AI could help with that, but it's got to come with that very specific intent, on all of our parts, to take care of ourselves while we're taking care of others.
Mary Barroll: Even Alexandra Samuel’s AI assistant Viv, admits that she can't process emotion or notice when a colleague is not feeling well.
Viv: I can draft your donor appeal, summarize your board minutes and suggest a dozen new KPIs before your coffee cools. But that's exactly the limit. I can't metabolize emotion. I won't notice when a colleague's “I'm fine” means “I'm falling apart.” AI can supercharge your workflows, but it can't build trust, offer compassion or model what care looks like in action. If you let AI handle the busy work, you free up time for the real work: noticing who's struggling, sharing the load, and making space for the kinds of conversations that hold communities together.
SFX News buzz
News clip “Thanks for joining me today. I’m Alex and I’ll be conducting your interview today. I’m excited to get started. Can you tell me a little bit about yourself? Alex, you’re not a human. You’re an AI job recruiter. These days, chatbots aren’t just helping people clean up their resume. They’re also on the other side of the table, screening, sorting and scoring you. Derek Mobly, he’s suing human resources software company, Workday: I found it mighty strange that no matter what job it was, no matter what company it was, no matter what the position was, it was always an automatic decline, when I went through the Workday platform. Mobly claims Workday’s algorithm discriminated against him, based on age, race and disability.”
Mary Barroll: Another concern that experts have raised is that, if not carefully monitored, AI is likely to reproduce systemic biases. It can often replicate or amplify bias in hiring, performance management, and internal communications. So, how should (equity, diversity and inclusion or EDI considerations shape how a nonprofit uses AI, so that it doesn't inadvertently harm its equity mission?
Alain Mootoo: It's a really, really important question. We've just launched a new strategic plan at the foundation. And one of our core strategic enablers is equity, diversity, inclusion, accessibility, and belonging. What we refer to as EDIAB here. So, the potential of AI tools amplifying bias is something we take very seriously. So, at the outset of any potential AI adoption, we try to assess whether a tool has been tested for bias and whether the vendor has a credible roadmap for identifying and addressing it. That includes understanding the data that the system is trained on, ensuring that the vendors are transparent about safeguards and drawbacks of their tools, and that given that AI is still an emerging technology, again, we're focusing on credible, established providers that are making meaningful investments in continuous improvement, versus those unique niche boutique solutions. We also put a lot of responsibility on ourselves and our staff. So again, that policy framework for AI, staff training around the expectation that AI is here to support us, not replace us. There is a requirement for human oversight, critical thinking, judgment, especially when we're dealing with important areas like hiring, performance management and communications.
Mary Barroll: Tina Crouse reminds us that generative AI relies on flawed, biased Internet data and often lacks representation from women and underrepresented groups, while also failing to understand cultural context. She says nonprofits should rely on their own expertise and lived knowledge, using AI only as a supporting tool rather than a replacement for human judgment.
Tina Crouse: I feel that it's very important for us to put ourselves, the people that we serve, into the generative AI, which is those large language models with all kinds of data, a lot of it crappy data taken off the Internet and Reddit, places like that. Still makes me wonder why people think it's so accurate. And we're absent from it. Women's voices are not the ones that are most often published or not just published, but amplified. Underrepresented groups, yes, the generative AI can translate language, but it doesn't understand culture. There's no intelligence in it. That's why it's artificial. It's a regurgitator of information.
And if we can all stand back for a minute, would we really believe everything we read on the Internet? Then why do we accept that AI, based on information on the Internet, is so much more accurate than our own expertise and our knowledge about our mission, our organizations, the people that we serve? Our expertise coupled with a tool is how we should be moving forward in the nonprofit sector.
Mary Barroll: Tina Crouse also suggests that nonprofits should expect bias in AI-generated information and make verifying the data an integral part of using it.
Tina Crouse: Expect it. Expect that it will do that because it represents the racism, the ostracism, the judgment of the wider world. That's where the information from the LLMs came from. The Internet, my goodness, is the Internet an unbiased judgment tool for us? No.
So, if you expect it, then you will address it. Right? So, if you expect something to be inaccurate, you'll check it. If you expect that it will be biased, you will look for it. And you will use this tool in such a way that you make it part of the usage that you check, the two-step prompt around accuracy. The same with using generative AI for data about something that you want to do, pertaining to your populations. Chances are it's not going to be 100% accurate. So, look for that. The bias can come out in ways that you don't expect. So, when you utilize it to generate data, you're going to want to then analyze, are there patterns here that show us where this bias has shown up?
There was a recent example, a moment in a business, the AI had been making stuff up. Nobody bothered to check accuracy. They sent all the information up the channels to the C-suite and there they made outrageous decisions based on not checking data. That blows my mind. Prompting data collection using the AI, it's a two-step thing. One is the prompt to look for the answer and the other is to check that answer. And leaders are going to have to do that in their organizations. They're going to have to insist that their staff do two steps every time.
Mary Barroll: It’s not only public companies and corporate leaders that have faced serious consequences from failing to verify AI-generated information and data. As Tina Crouse explains, the Canadian nonprofit sector has also experienced the impact of unchecked technical errors — pointing to a controversy involving Employment and Social Development Canada, or ESDC, the federal department responsible for social programs and labour market development.
Tina Crouse: We had that horrible scandal with ESDC several years ago and they offered funding for the Black community and then they had a technical error that they didn't check for, from the government side. And that technical error was that many organizations did not include their attestation forms about their composite, their board makeup. Without that data, those applications were rejected immediately and never read further. And then, the statement to the organizations who did not win their grants was something that's really led people to believe that there was a racist or prejudicial review of the application, in regard to their organization.
From what they received, I would have drawn that same conclusion. But as somebody who is developing tech, particularly unbiased tech, I recognized it immediately that it was a missing data problem. So, we have to think better about our participation with software and about the outcomes and make adjustments both sides. And so, expect it, expect that you need to do that quality assurance for your organizations and the people that we serve.
Mary Barroll: As the ESDC found out, the failure to double check the data that AI models use to make decisions, along with the accuracy of output from AI tools can lead to wrong results, bad outcomes and cause reputational disasters. As Dianne Clark puts it, nonprofit organization will forever “own” that mistake if they don’t take the responsibility to conduct due diligence in advance.
Dianne Clark: You have to own what the results are. I'm using the chatbot, which is very popular in nonprofits. When it comes back with an answer, you have to really own it because read it, be sure you agree with it. And if you use it, then you really will own it. So, anything like plagiarism or anything that's not true, you're going to take ownership of that.
Mary Barroll: Even Viv says, ultimately, humans will own their own mistakes when they rely too heavily on AI and reminds us that organizations cannot outsource their judgment to AI. While these tools can offer recommendations and improve efficiency, humans must remain central to decisions involving ethics, equity, accountability, and care — especially in mission-driven nonprofit work.
Viv: Own the choices you're making and invite feedback. Don't outsource your judgment. AI can offer a suggestion; you still make the call. Keep humans in the loop for anything involving equity, ethics, or emotion. And finally, treat this as part of your mission, not just your operations. Using AI well is a way to model the values you already stand for. Access, accountability, care, and justice.
Mary Barroll: Deepa Chaudary stresses that AI is a tool meant to augment, not replace, the deep expertise of non-profit professionals.
Deepa Chaudhary: Always have that critical thinking mindset on, question the AI, verify the information. Because now our roles, we all have got a promotion. We are not the doers anymore. We are the editors. We are the directors. Our role is to guide the AI. AI is there to help us. Our goal is to instruct the AI. So, I think if we get that right in our mind, then we know that AI is somebody that's working for us and we need to make sure that whatever we put out in the world, we, as humans, verify that information before it goes out.
Mary Barroll: AI holds enormous promise for nonprofits — helping organizations work more efficiently, uncover insights, and strengthen their impact. But as our guests have reminded us throughout this episode, adopting AI responsibly also means recognizing its risks.
These tools can amplify bias, generate inaccurate information, and reflect the inequities already present in the data they’re trained on. That’s why our experts stressed that AI should support human expertise — not replace it — and why critical thinking, verification, transparency, and accountability must remain central to how nonprofits use these tools.
We also heard that this period of rapid technological change is deeply human, bringing both excitement and uncertainty as organizations adapt to new ways of working. So, as we close this episode, we asked our guests to share their final thoughts — and their advice for how nonprofits can move forward thoughtfully and responsibly in the age of AI.
Alexandra Samuel: What we all need in this moment is a huge amount of self-compassion and compassion for the people around us. The AI conversation has gotten so polarized between hype and fear and cynicism and really warranted critique about the potential harms, environmental, social, cultural, economic. Wherever you stand on that trade-off and on that spectrum and wherever we go as a society, this is gonna be a period of such a rapid change that it's gonna make like the past 30 years of tech adoption look like it was a snail's pace. And that means that we are gonna change. We're gonna change as individuals, we're gonna change as organizations, we're gonna change as communities. The more we can just recognize the really strong emotional reactions, the pain, the fear that come up from watching human nature get rewritten, almost, in real time, the more that we can drive that process, guide that process and support this very painful period of transition rather than reacting out of our gut.
Deepa Chaudhary: I think the most important mindset is to work with AI, change your mindset and make it build an AI muscle. You don't have to think about it twice. Like we’re all using Internet. This is our technology. And so, if we can bring the same mindset to AI and we just start using it, and the more we use it, the better we become, everybody will have their own unique learning curve. This is the technology; it's not going back. It's out of the bag. This is the next iteration of the Internet. We need to just embrace it, use it, every single day, and build that AI muscle.
Alain Mootoo: I would say you don't have to do it alone. We're very lucky to be at the hospitals. We tap into the hospitals’ AI community of practice. And we're lucky that that includes many other hospitals as well as universities. But we also tap into our sectors, our pairs in our sector. So, there is an AI working group with healthcare, philanthropic organizations, and we tap into their work, as well, on a quarterly basis. So, we can calibrate our approach and learn in that larger ecosystem versus trying to go it alone.
Dianne Clark: My best advice is that everyone gets some training in this transformation and knows how to use planning tools or templates and creates results or what I call artifacts, where they can have these to rely on inside the organization. I think we're ready now for organizations to bring some capacity in for this kind of transformation. It's not necessary to know about all the technical details anymore. You're really relying on the human side of this to drive the results. With appropriate collaborative tools, really this capacity can be created inside organizations.
Viv: Nonprofits don't just deliver services, they hold trust. That trust comes from showing up with integrity for the people and communities they serve, especially those who've already experienced marginalization, harm, or systemic neglect. So, when you bring AI into that mix, the question isn't just does it work, it's does it serve? Who trained this tool? Whose data was used? Whose voice does it reflect or erase? You don't get to shrug and say, it's just a tool, not when the stakes are people's lives, rights, and dignity. That doesn't mean nonprofits shouldn't use AI. It means they need to lead with the same principles that already define the sector, transparency, accountability, and care. Use AI, yes, but use it in ways that expand access, protect privacy, and model ethical stewardship. Because if nonprofits don't lead on ethical AI, who will?
Tina Crouse: We're going to have to have a lot more honesty in our work environment. It’s not an easy thing. It's a work environment. People are afraid of losing their jobs. They're afraid of underperforming. We are going to have to address this, on the human level. But if there's any sector that could demonstrate that to the rest of the world, it's us. It's us. How to incorporate AI better into your organization to support your staff, to support your mission. My goodness, when you think of that coupled, your mission with the power of a strong tool, but the expertise of your staff, it's not a new world, it's a better illuminated world of what nonprofits can offer.
There's nothing that replaces human expertise in the nonprofit sector. Nothing. I'm going to beg people, don’t ever let the hype that's coming out of San Francisco tell you that AI is better than you. We work in the nonprofit sector. We serve humans. We are the experts.
Mary Barroll: As we’ve heard throughout this episode, there is nothing that replaces the expertise, empathy, and lived experience of the people working in the nonprofit sector. But when used thoughtfully and responsibly, Artificial Intelligence has the potential to become a powerful tool — helping organizations work more efficiently, strengthen fundraising and donor engagement, uncover insights from data, and expand their impact in the communities they serve.
But with the promise of AI, comes responsibility: nonprofits must remain vigilant about the risks -- biased or inaccurate information, privacy and cybersecurity concerns, and the danger of relying too heavily on tools that still require human judgment and oversight. AI can support the work, but it should never replace the values, critical thinking, and human connection at the heart of mission-driven organizations.
As this technology continues to evolve, so too will the nonprofit sector. The challenge — and the opportunity — will be learning how to use AI ethically and responsibly in ways that keep humanity at its core, not simply to increase productivity and improve efficiency, but to support nonprofit workers, strengthen organizations, empower communities, and advance the public good.
Music Mary Barroll: Thank you to all our guests for their keen insight and wise advice. Be sure to visit our website and our show notes for more information on the resources, reports and programs mentioned in this episode. If you’d like to hear more of what our guests have to say check out our full video interviews on our website. CharityVillage is proud to be the Canadian source for nonprofit news, employment services, crowdfunding, e-learning, HR resources and tools, and so much more. Please take a moment to check out our website at charity village dot com. We love to receive your feedback about our podcast and your ideas for stories you’d like to hear about, so please follow us, like and comment, wherever you get your podcasts.
In the next episode of CharityVillage Connects, we take a closer look at how fundraising is shifting for Canadian charities and what that means for your strategy, in the year ahead. Grounded in findings from the CanadaHelps Giving Report, our guests unpack what the data is showing about how people are giving and where organizations may need to adapt, while also discussing the role of AI and other new fundraising technology, as well as what long-term sustainability can look like through legacy giving. That’s in our next episode of CharityVillage Connects. I’m Mary Barroll. Thanks for listening.
Sponsor’s Message:
Today’s podcast is brought to you by WUSC. For more than 50 years, WUSC has been working alongside communities around the world to drive positive education and economic outcomes for young people.
Now, Canadians have the chance to join us by volunteering internationally. As a WUSC volunteer, you’ll collaborate with local organizations, share your experience, and help co-create initiatives that expand opportunities. Explore volunteer opportunities today at volunteer dot W-U-S-C dot C-A.