In our recent CharityVillage Connects podcast episode, Responsible AI Adoption for Nonprofits, we gathered technology and nonprofit sector experts together for an important discussion on AI use.
As part of this podcast episode, we were able to speak with Alain Mootoo, Chief Operating Officer of CAMH Foundation. Alain shared with us how CAMH Foundation is approaching the implementation of AI tools. We’re pleased to present an excerpt of his interview here.
We started off by asking Alain how nonprofit organizations might embrace a responsible approach to AI that isn’t solely driven by efficiency gains.
Alain Mootoo: 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. So 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 labour-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. And at this stage of our journey, we adopted ChatGPT 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. 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.
Alain went on to describe a strong, lower-risk use case for using AI, as well as commenting on choosing AI vendors.
Alain Mootoo: When it comes to thinking about use cases, we really go back to sort of the most strategic level. So why are we thinking about AI?
We’re a fundraising organization. So ultimately, we’re here to raise more money for our cause. When we think about AI, we start by asking, where is fundraising not being fully optimized? And where are our teams slowed down by inefficiencies? Our vision with AI is to equip our fundraisers with more actionable information and tools so they can raise more funds, build stronger relationships, and create better experiences for our donors. A big part of that is improving how we understand our donor communities, who they are today and who they might be in the future and how we design meaningful journeys and experiences for them. Fundraising is also very complex, so donors’ priorities are evolving, research and care breakthroughs are moving quickly, and all of this is happening in a very dynamic political, social, economic, and cultural context.
So for us, when we think about AI use cases, we think about those that will help our teams move into action faster and navigate complexity more efficiently and effectively. Practically speaking, AI use cases for us are the ones that are focused on helping synthesize large amounts of information, large amounts of data. They help us draft briefings and communications, they help us analyze publicly available insights on our donor interests, and they improve how we design and document our workflows.
Like many organizations, we also have to reflect on whether we are going to build or we’re going to buy AI solutions. And for us, it’s very much a buy decision. Our role as a charity is to not develop the AI technologies ourselves, but to thoughtfully adopt credible tools that strengthen how we work.
When we think about the AI environment, it’s not regulated in a very mature way. It’s still evolving, especially here in Canada. So we’re being very careful about choosing AI vendors who are most likely to have the highest standards and the most robust security and quality programs. We stay away from those boutique new providers that may not be paying as much attention to quality and privacy.
Alain filled us in on another risk related to AI use within organizations – shadow AI, where staff are using public AI tools outside of the organization’s knowledge or policies, potentially opening up the organization to cybersecurity risks.
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 ChatGPT licenses for every staff member. We also upgraded Teams licenses to enable AI note-taking. 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. But 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 train 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.
From a cultural perspective, we’re also trying to create that space for experimentation. We have monthly staff meetings where we talk about our AI journey. We have lunch and learns where staff can talk about their work, including how they’re using AI. We’re hoping all of that really creates the environment where people feel encouraged to innovate, and they’re less afraid of sort of change and trying new tools.
Alain also offered some thoughts on some potential drawbacks for staff using AI, including deskilling or increased cognitive strain.
Alain Mootoo: We certainly have heard about cybersecurity and de-skilling, definitely. What we’ve also really been hearing a lot about is concerns about shifting workload, productivity and cognitive overload. So it’s very interesting that, you know, as AI speeds up the production of outputs – so more content, more analysis, more data – human beings have to review and verify that data and govern it. So in some cases, that oversight responsibility can create an increase in mental effort if there isn’t some thoughtful implementation.
If we adopt multiple AI systems too quickly, that oversight responsibility can create an increase in mental health risks, and 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 can contribute to fatigue and exhaustion.
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.
Want to hear more from Alain Mootoo? Listen to his full interview in the video below.
Listen to Alain Mootoo and other sector experts discuss how nonprofit organizations can responsibly embrace AI adoption in our new podcast episode. Click here to listen.
The views expressed in this article are the author’s alone and do not necessarily represent those of CharityVillage.com or any other individual or entity with whom the authors or website may be affiliated. CharityVillage.com is not liable for any content that may be considered offensive, inappropriate, defamatory, or inaccurate or in breach of third-party rights of privacy, copyright, or trademark.
