I’ve had many conversations with EDs and nonprofit staff about why nonprofits aren’t using AI, even when some of the benefits seem obvious.
The idea of implementing AI into your work is compelling: streamline operations, improve donor engagement, stretch limited resources further…
But why aren’t more nonprofits adopting it? The reason isn’t lack of interest or resistance to innovation. Rather, it’s all about the money.
The Margin Problem
A big part of why nonprofits aren’t using AI comes down to funding. According to Scott Brighton, CEO of Bonterra, a software company serving nonprofits, a healthy for-profit company might run at 40% profit margins. But in the nonprofit world, organizations give away roughly 80% of every dollar that comes in. This leaves very little to invest in internal infrastructure, technology, or staff development.
Most nonprofits are small, with annual revenue under $10 million. They don’t have in-house IT teams and instead rely on outside vendors for software and tech support, which gets expensive fast. For organizations already stretched thin, adding AI tools to the budget is usually a matter of whether they can afford to invest in technology at all.
I’ve seen this pattern consistently. Funders and individual donors want to fund the “warm and fuzzy” projects, like scholarships, food and coat drives, and community events. Behind-the-scenes technology investments aren’t sexy. A new CRM system or AI tools don’t typically get a funder’s name on a plaque. But, without that infrastructure, staff can continually waste time on manual processes.
Many boards reinforce this with a “nonprofits should look poor” mentality, even if you’re running a six-figure or multi-million-dollar operation. Just because you’re not generating profits for shareholders doesn’t mean you shouldn’t invest in the tools your staff needs.
For example, I’ve seen development directors spend hours stuffing envelopes instead of building donor relationships. Sure, outsourcing to a local print shop costs money. But so does paying a development professional to do clerical work, while also losing the revenue they could have generated by taking them away from donor cultivation. Nonprofits see the invoice from the print shop and think they’re saving money by doing it in-house, but they’re not calculating the actual cost of staff time plus lost opportunity.
The Skills Gap No One’s Talking About
Even if an organization can afford AI tools, many nonprofit staff haven’t been provided the foundational technology training needed to use them effectively. Frontline staff at many nonprofits are already doing multiple jobs with limited resources. Asking them to learn and implement AI tools without investing in training sets them up to fail.
In my work, I’ve seen organizations struggle not because they lack technology, but because they haven’t been trained to use it effectively. They’re running HR functions through spreadsheets, tracking expenses with Word documents, and building everything from scratch instead of using systems designed for these purposes.
The argument is always about cost. But again, that calculation ignores the hidden costs: staff time building templates from scratch, the lack of automated alerts, the loss of institutional knowledge when an employee leaves and takes their homemade system with them.
Before considering AI, we need to talk about whether organizations have foundational systems in place and whether staff know how to use them.
What Actually Works: Targeted, Mission-Aligned Use Cases
There are nonprofits making AI work.
Bonterra recently introduced “agentic AI” tools designed specifically for nonprofits. Instead of requiring massive upfront investment, these tools support staff by handling specific tasks like donor segmentation, freeing up staff time for relationship-building and program work. Focused tools like these help address why nonprofits aren’t using AI at scale: they solve a real problem without overwhelming staff.
Tech Goes Home deployed a chatbot on its website to answer questions about services, which is saving staff time while also helping users find information. And because Tech Goes Home focuses on digital equity, their AI work includes teaching their community how to use the technology.
These examples share a common thread: they’re starting with one or two strategic and high-impact use cases that solve real problems and align with organizational capacity.
How Nonprofit Leaders Can Get Started
If you’re trying to figure out where AI can fit within your organization, here’s where to start:
Assess your digital readiness first. Before you invest in AI tools, make sure your team can use your current systems effectively. If not, that’s where your investment should go.
Pick one use case and involve your team. Don’t try to map out a comprehensive AI strategy on your own. Pick one small, repeatable pain point and experiment. Ask your staff where the bottlenecks are and where a tool could help.
Push back on unfunded mandates. If funders or board members are excited about AI, that excitement needs to come with investment in infrastructure and capacity-building.
Avoid the one-sentence prompt trap. The biggest mistake I see is someone typing “write me a fundraising email,” getting back something generic, and concluding that AI is junk. AI requires context. The more specific you are about your organization, your audience, and your constraints, the more useful the output will be. AI is a tool that makes work easier when used well. It’s not a replacement for human judgment.
The Bottom Line
AI can help nonprofits work smarter, but only if organizations have the infrastructure, skills, and realistic expectations to make it work.
The barriers are real: tight budgets, limited technical skills, and a funding culture that prioritizes programs over backend systems. But they’re not insurmountable.
Start small. Pick one task that’s eating up staff time. Make sure your team has foundational technology skills. Use AI as a thought partner that requires context and oversight, not as a magic button that works on command.
If you’re a funder or board member excited about AI, help fund the infrastructure and training that make adoption possible.
If you’re ready to explore AI strategically, ask yourself: What’s one operational headache costing us time every week? That’s your starting point.
This post was inspired by reporting from Eoin Higgins in IT Brew.
Image by cocoandwifi from Pixabay.