I'm writing this from the Hyatt in Seattle, still processing day one so far of Microsoft's Global Nonprofit Leaders Summit. Forty countries represented. Big names on stage. Brad Smith comparing AI to electricity. Nicole Herskowitz showing how Copilot can turn a quarterly report into a polished donor presentation in minutes.
And sitting in my seat, I felt two things at once: genuine, geek-level excitement, and...stress. Stress from the nonprofit barriers that are ever-present.
Because here's the thing. I'm one of the AI enthusiasts (surprise, surprise). I want to go fast and break things (I do that a lot). I want to redesign our entire organizational infrastructure and processes to work better and more efficiently (I am annoyingly persistent about this, ask my colleagues). I've spent eight years in nonprofit, five at Zero Abuse Project, pushing for exactly this kind of transformation.
But I also know, from those same eight years, that nonprofits don't move like tech companies. They can't. And watching this summit, I kept wondering: do the people on that stage understand what it actually looks like on the ground?
The Numbers That Should Make Us Optimistic
Let me be clear. The vision Microsoft laid out isn't wrong.
Brad Smith talked about AI "diffusion" being the key to societal benefit. He announced $50 billion to bring AI to the Global South. He told a story about a history teacher in Delhi who used Copilot to create a full stage play his students performed, something a theater expert had said was impossible.
Justin Spelhaug, President of Microsoft Elevate, shared case studies that made my heart sing. Children's International is saving 5,000 hours per month across 500 staff members using AI agents for translation. UN Human Rights is using AI to transform raw field data into decision-ready briefings in minutes. These aren't hypotheticals. These are real organizations doing real work.
And they announced Microsoft Elevate for Changemakers: an "AI for Nonprofits" professional certificate, live training, a fellowship program. Real infrastructure to help nonprofits build capacity.
I want all of this to happen. I believe it can happen.
The Reality Check
But here's what the summit didn't spend much time on.
A recent benchmark study of 346 nonprofits found that while 92% use AI, only 7% report major improvements in organizational capability. 81% use AI individually without shared workflows. 47% have no AI governance policy. PR Newswire We're experimenting, not transforming.
Some organizations are charging ahead; some are cautiously experimenting; some are overwhelmed. Many simply don't have the knowledge, infrastructure, or funding to explore AI in a meaningful way. Bridgespan
Many organizations with annual budgets under $500,000 are hesitant to implement AI due to financial constraints, lack of technical expertise, and uncertainty about where to start. TechSoup
Larger nonprofits with annual budgets exceeding $1 million are adopting AI tools at nearly twice the rate of smaller organizations (66% vs. 34%), underscoring a growing digital divide. NonProfit PRO
And that's before we talk about the political climate.
One in three US nonprofits that serve communities lost government funding in early 2025. The Conversation. Specifically, 21% reported losing some government funding, 27% experienced a delay, pause or freeze, and 6% received a stop-work order. Real Economy
The 43-day government shutdown in October 2025 was the longest in U.S. history The Nonprofit Alliance, and due to the 2025 funding cuts and the compounding nature of additional service gaps that nonprofits filled during the shutdown, additional funding challenges are anticipated for 2026, which may include staffing and service cuts. The Nonprofit Alliance
Many nonprofit leaders express concern about how the political climate may affect their funding. In response to actions taken by the federal government this year, leaders say that "we are actively seeking measures to untangle ourselves from public funding," but "we simply do not know whether private foundation grants will keep up." As one leader states, "Maintaining clarity of mission and financial sustainability over the next few years, in the current political climate, will be an existential challenge." Cep.org
How do you tell an organization that just lost a third of its federal funding, that's running on a skeleton crew, that's worried about making payroll next month, that they should be excited about AI transformation?
The Stat That Made My Eyes Roll to the Back of My Head
During the session on building an inclusive AI economy, Justin Spelhaug shared data that I can't stop thinking about.
Analysis from the Washington Post and Brookings shows that 6.1 million workers face both high AI exposure and low adaptive capacity. These workers tend to be concentrated in clerical and administrative roles, and about 86% are women. Brookings
Eighty-six percent, folks.
Secretaries were among the 6.1 million largely clerical and administrative workers considered both highly exposed to AI and with the lowest estimated adaptability. Women make up about 86 percent of those most vulnerable workers, suggesting the negative effects of automation won't be borne equally across society. The Washington Post
Justin argued that nonprofits are uniquely positioned to address this challenge. Proximity to communities. Trusted voice. Ability to be a delivery layer for AI skilling.
I want to believe that. I do believe that, in theory. But I also know that the same nonprofits being asked to become the AI skilling delivery layer are the ones burning out their staff, losing federal funding, and trying to figure out how to keep the lights on.
Nearly 90 percent of nonprofit leaders report some level of concern about their own burnout. A similar percentage note that burnout is affecting their staff. Cep
As one leader put it: "Staff at social and racial justice organizations are tired. The intensity of trying to protect communities during COVID rolled into trying to protect those communities from a rising right-wing backlash. This goes beyond burnout. It is a psychic and physical drain on folks' ability to function." Cep
What I Actually Think
I'm not writing this to be cynical. I'm writing it because I think the optimism gap matters.
If we pretend that nonprofit AI transformation is just a matter of will, training, and the right tools, we're going to leave behind the exact organizations that need it most. The small ones. The underfunded ones. The ones serving communities that don't have tech champions on staff.
And if we don't pay attention to who gets displaced by AI, we risk creating exactly the kind of harm that nonprofits exist to prevent. Mark Muro, a senior fellow at Brookings, said the most vulnerable workers "may be out of sight and out of mind" to policymakers and the American public. The Washington Post
Here's what I took away from day one:
The technology is real. The potential is real. Microsoft is putting serious resources behind nonprofit adoption, and that matters.
But diffusion isn't automatic. And the people in this room, the forty countries of nonprofit leaders, are going to go back to organizations that are stretched thin, politically targeted, and exhausted.
The question isn't whether AI can help nonprofits. It's whether we build the support systems that make adoption actually possible for the organizations that need it most.
And whether we pay attention to the 6 million workers, mostly women, who may get left behind while we're celebrating the transformation.
What I'm Watching Tomorrow
Day two features Melinda French Gates and Trevor Noah. I'm curious whether the conversation will get more real about the challenges, or whether it will stay in the optimism lane.
I'll report back.
In the meantime, I'd love to hear from other nonprofit leaders. Does this match your experience? Are you feeling the optimism, the overwhelm, or both?
