TL;DR: We tend to talk about AI adoption as a technology challenge with a human dimension. I think it’s actually a leadership development opportunity with a technology dimension. Three days at the Nonprofit Technology Conference, sponsored by NTEN, sharpened that instinct and left me with a clearer sense of what it looks like to navigate this moment with intention, agency, and the systems awareness the future is going to require.
During the first term of this administration, I remember likening the news to a ‘firehose of atrocities’. Now, a pandemic, multiple wars, and a second ongoing administration compounded with the existential threats of AI have made that firehose look and feel like a sprinkler.
Everyone is tired. And we’re tired in so many ways. But, in the nonprofit sector, I see a deeper layer of that exhaustion. The kind that comes from feeling like what you’re doing might not amount to anything, like the ground keeps shifting, like the systems you’re operating in are working in directions you didn’t choose and can’t quite see. This makes it so incredibly challenging to meet this AI moment with imagination, agency, and proactivity.
Anil Dash named it in his opening keynote at NTEN’s Nonprofit Technology Conference last month in Detroit: burnout in this sector isn’t primarily about hard work. It comes from moral injury — the persistent, grinding sense that our efforts will be undermined, that the forces shaping the environment we work in are larger and less aligned with our values than we thought.
I’ve been mulling over that framing since I came back from Detroit, because I think it connects directly to something I see in almost every organization I work with right now. The challenge of AI adoption isn’t primarily technical. It’s not even primarily about change management in the typical sense of linear implementation.
It’s about trying to do mission-driven work with intention and integrity inside systems (and sometimes organizations) that are generating that moral injury — systems that are moving fast, rewarding speed over reflection, and not particularly designed with our communities in mind. It’s about operating in narratives that center only hype or doom, thus limiting agency. It’s about our organizational AI conversations being hyper-focused on “efficiency” and thus missing any conversation of what is potentially transformative for our work. These are all deaths by a thousand cuts that make stewarding anything relating to AI an incredibly challenging task. [Soapbox side-rant: stop putting AI solely in IT departments, just stop!]
That’s a hard context to navigate. But I think how we navigate — the agency we build or suppress, the awareness we develop or ignore, the moves we make with the power we actually have — is one of the most important leadership development opportunities this sector has seen in a long time. Yes, there are certain leadership skills that will best serve this moment, but what if the moment itself is the opportunity to build those skills across a broader group of people?
What if we approached AI in that way?
Here’s what I mean. When we talk about AI adoption in the social sector, we tend to ask: how do we help people get comfortable using these tools? How do we manage the transition? How do we develop the policies and protocols that keep us responsible? Or, more annoyingly, how do we use AI to be more ‘efficient’ (which then quickly just becomes cramming more into our days)?
Those are real questions. They’re worth asking during the process. But they can’t be the questions because they position AI as a thing happening to individuals and organizations, with leaders as managers of the response. And I think that framing undersells both the challenge and the opportunity. It puts us on a reactive back foot, not leading with a vision of “what does this make possible”?
As someone well-versed in principles of Emergent Learning, I decided to try to bring some framing questions with me to NTC that I thought my be more galvanizing for my own work and help me learn from the conference in deeper ways. I've been reading about plant intelligence lately (more on that later), and I kept thinking about the concept of the “overstory” — the canopy layer whose conditions shape everything below it. That's what I was looking for at NTC. Not just session-by-session takeaways, but the pattern above them.
I decided to explore: What does it take for nonprofits to adopt technology in ways that actually serve their mission — not just technically, but in practice, in relationship, over time, and under pressure? How are organizations making sense of AI right now, and what does that tell us about what they actually need? What makes AI implementation and adoption feel humanizing rather than dehumanizing? And, what does “good” AI use look like, and who gets to decide?
What I realized is that ultimately, these are leadership questions. And the answers have everything to do with power, trust, and the ability to see the systems you’re operating in. To continue the metaphor, leadership is the practice of working at the canopy level. Of tending to the overstory, the conditions that shape everything underneath, so that what emerges in the understory (the tools, policies, trainings, individual choices/decisions, etc.) becomes rooted in agency, rather than moral injury.
If we reframe AI and a leadership opportunity, we can then see that the most sustainable, effective, long-term hope for our leadership is to allow AI to invite (or push) us to develop the capacity to work at this overstory level. To stop micromanaging the understory and start asking what conditions we’re creating.
But, what does that actually look like in practice?
I’ve long believed that the same systems thinking lens that serves culture change management and DEI work is the most relevant skillset for this AI moment. NTC reaffirmed and deepened that instinct.
In the keynote, Anil Dash’s power lens challenge wasn’t just abstract. He put it up on the screen. Everybody has some amount of power. We don’t always see it, just as we often can’t see the systems we’re embedded in. As he pointed out, “The purpose of the system is what it does.” And that’s not just a cynical statement, it’s a diagnostic one. It asks us to look at outcomes rather than intentions when evaluating whether something is actually working. And it asks us to notice when we’re lending our power to systems that are generating that moral injury rather than relieving it.
Another session that I attended on Advancing the Black Tech Agenda put the sharpest version of this on the table: making the motivations of tech companies, policy writers, and lawmakers transparent is itself an intervention. “Sunlight” does a lot to expose these mechanisms and how they work. That applies at the organizational level too — when we name what we’re optimizing for and why, we make it possible to notice when we’re drifting.
The next day’s keynote on Public Interest Technology named the drift pattern directly: when tools don’t fit the mission, the mission tends toward the tools — incrementally, through a hundred small accommodations, each one reasonable, until you look up and realize your work has reorganized itself around your software. No single decision caused it. It happened in the gap between intention and attention.
The leadership practice here is developing the capacity to see those systems and that drift — not once, during a policy-writing retreat, but as an ongoing habit of attention. Mission alignment isn’t a design-phase checkbox. It’s a practice. It’s an ongoing, imperfect, dynamic practice because we are social beings existing in systems that are always in flux.
Seeing this habit of attention as a practice is on that invites agency. It’s a small shift, yes, but simply giving permission and showing where agency exists at this moment can be incredibly powerful when everything feels so out of control
I saw a version of this take hold recently in a kickoff session I facilitated for the Colorado Collaboratory on Equitable Inquiry programming arc on AI. One of the key takeaways participants named was: “I have agency when it comes to AI.” I didn’t realize it, but I had designed toward that outcome without fully naming it. Anil’s framing helped me understand why it felt so important to get there. Agency — the felt sense that you are a conscious actor in this, not just a recipient of it, and that you can make choices — is a leadership outcome. And it has to be explicitly cultivated, because the ambient culture is working hard against it.
Across panels at NTC, the same thread surfaced in different forms: fear is the biggest obstacle to technology adoption right now. Not just confusion, not just lack of access. But fear. Fear of clicking the wrong thing, of looking incompetent, of being judged unfairly, of doing harm accidentally, of being left behind.
What moves fear isn’t more explanation. It’s guided practice. Peer support (a digital inclusion panel reminded us that learners will go to a peer before an instructor — always). Spaces where it’s genuinely okay not to understand yet. Positive reinforcement for trying new things, even if they don’t work out. A growth mindset frame that asks “how do I improve?” rather than landing on “I can’t do this.”
These are, notably, not AI interventions. They’re human ones. They are classic social learning theory in that they’re the conditions that have always been required for learning to happen, and they’re the conditions that AI adoption often skips in its hurry to get to the tools and the ROI.
Moreover, understanding this led me to consider another important lesson for leaders: the way you’re feeling right now is a moment of empathy for the people you serve. If you feel overwhelmed by the pace and the options and the fear of doing it wrong, you have direct access to something important. Then you now have a glimpse of how it feels to navigate systems that weren’t designed for you. That’s not a weakness in you as a leader. It’s data. And, if you decide to listen to it then it could be the beginning of building something that actually fits for your organization.
Building trust — for your team, within your organization, inside yourself — is leadership work. It has to come before tools. And building it, under these conditions, is some of the most demanding and necessary leadership practice available right now.[For a deeper dive on the paradox of trust in AI leadership, see this new piece I co-authored with Micela Leis for CCL!]
A TAG session on grant reporting in the age of AI surfaced another powerful observation: those who define the problems in our sector end up defining the solutions. For example, if funders define the problem of grant reporting, they’ll build solutions that serve funders. Instead, the people experiencing the problem need to be designing the solutions. I found myself sitting with what that observation means beyond grant reporting.
If AI literacy/fluency programs are designed primarily by consultants (a category I am in) without meaningful input from the people doing the learning, we may be reproducing the pattern we’re trying to address. If AI policies are written by leadership teams without the voices of the staff implementing them, the same thing happens. Who is in the room when the problem gets defined will shape what solutions are even imaginable.
Asking questions — who defined this? and whose voice is missing? — is a leadership practice. And, to connect to my earlier point above about what DEI can teach us in this moment, it’s a practice rooted in equity and systems thinking. It requires the willingness to slow down at the exact moment when the culture is pressuring you to speed up. Or, in the better words of Dr. Bayo Akomolafe: “The times are urgent: let us slow down.” Practicing that slowness inside of a chaotic AI moment that appears to be rewarding speed, efficiency, and ‘optimization’ is an intentional leadership choice.
These systems-level questions connect directly to something I wrote about in my last post: cognitive protection — the individual-level work that has to accompany any serious AI adoption effort.
It’s difficult to exercise agency in a system you can’t see. And you can’t protect your own thinking without understanding the design incentives of the tools you’re using. AI tools are software products built by companies with revenue models, and those models don’t always align with your wellbeing or your judgment. In fact, they are often intended to exploit it.
The same patterns we learned from social media — attention erosion, dependency loops, the engineering of engagement over reflection — are present here too, with the added complexity that we’re interacting with these tools conversationally, in ways that touch our sense of self and our ways of thinking.
Building awareness at the individual level and building awareness at the systems level are pieces of the same work, just at different scales. Moving between them by noticing what’s happening in your own thinking while also seeing the systems that are shaping that thinking is a practice. I’d also argue it is one of the most transferable skills this moment is offering us. Leaders who develop it won't just navigate AI better. They'll be more equipped for whatever comes next.
Systems are large and power is diffuse. Most of us don’t have levers that move everything at once. What we do have is influence within our sphere, accumulated over time, in relationship with others doing the same or adjacent things. One of my clients said something during discovery work that I keep coming back to: “we operate within systems we don’t control.”
That’s true. And it’s also not the end of the story.
Recognizing our connections within those systems, tending them intentionally, and making the most strategic moves available to us with the power we actually have — that’s how conditions shift. Not through a single dramatic intervention, but through many actors making many small, intentional moves, in enough coordination to change the shape of things. (Related, I’ve been obsessed with reading about plants and trees and what plant intelligence can teach us about leading in this moment. If you haven’t read The Light Eaters yet, then I highly recommend it!)
So the question I’d invite you to sit with is this: given the system you’re operating in and the power you actually have, what’s the most strategic move available to you right now? It may not look or feel like leadership in the class, traditional, hierarchical, patriarchal sense, but I assure you that it absolutely is leadership and it’s a moment of practice.
It might be attending a local zoning meeting where data center placement is on the agenda — because small and midsize towns are already making decisions about tech infrastructure that will shape your community for decades. It might be redesigning a training to start with values and trust conditions before tool features. It might be asking, before the next AI decision gets made, who defined this problem and whose solution this will be.
Or it might be quieter than that. Naming agency as an explicit outcome in your next session. And then designing for it. Creating space for your team to say what they’re actually afraid of. Protecting your own analog life in whatever small ways you can manage this week.
None of these moves is sufficient on its own. But that’s the point. All of them are worth making. And making them, I think, is what leadership looks like right now — not “managing” an AI transition, but developing the capacity to act with intention inside systems that are moving fast, that weren’t built for us, and that we are nonetheless not powerless within.
That’s the leadership training. We’re in it.
What question are you sitting with about you or your organization’s relationship to AI right now? I’d love to hear what’s alive for you.
Sharing is caring, especially when it comes to competing in the algorithms! If you liked this post, please consider sending it to someone else (and then have a conversation about it!)
💡Did you know?🌱 Helping mission-driven leaders, teams, and organizations thoughtfully and responsibly adopt AI is what I do! If you want to learn more, please reach out!
To stay up-to-date on my writings across platforms, please join my monthly mailing list (no spam, ever!). Free, downloadable guides and case studies coming soon!
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.