AI Is Already Inside Your Organization. Do You Know What To Do With It?

Created By: Lauren Jones Kenny
June 25, 2026

Over the course of two executive convenings in Miami and West Palm Beach, a single theme emerged around the adoption of artificial intelligence: nonprofit leaders are no longer debating whether AI belongs in their organizations, but are genuinely uncertain about what to do next and how to keep up with the rapid rate of change.

The leaders in these rooms represented a cross-section of the South Florida nonprofit sector, varying in size, mission area, and budget, and their experiences with AI varied accordingly. However, the underlying pattern was remarkably consistent. AI arrived in their organizations before any formal decision had been made about it. And in nearly every case, leadership was operating without the governance infrastructure to direct it.

The gap between the pace of adoption and the pace of institutional response is the defining challenge facing the nonprofit sector on AI right now. And the data from our own research bears it out.

What the Numbers Reveal

Across the organizations we work with and the broader sector research we track, several findings stand out.

92% of nonprofits report using AI in some form. Based on what surfaces in facilitated conversations with employees, the real figure, when you factor in AI-assisted tools in use across search engines, conferencing tools, and email inboxes, is 100%.

While 30% of nonprofits using AI have reported meaningful fundraising gains (a number that sounds encouraging until it is placed against the adoption rate), our research has found that only 7% of nonprofits report major mission-level improvement from AI, and only 5% of AI initiatives reach their stated goals. Most organizations using AI are not yet converting that use into measurable impact in mission fulfillment or on the bottom line.

The explanation is not technological; it is organizational. While the tools available to nonprofit organizations today can produce significant results, 76% of nonprofits lack any formal AI policies. Without them, AI use inside an organization is ungoverned, unaccountable, and, despite the best intentions of the staff driving it, largely directionless.

A Sector Moving Faster Than Its Leadership Structures

One of the more striking observations from these convenings was how clearly leaders recognized the gap between where their organizations were and where they needed to be, and how infrequently that recognition had translated into action.

When asked to place their organizations on a five-stage maturity model mapping the nonprofit AI adoption journey, many leaders self-identified at Stage 1 or Stage 2. This was not due to a lack of familiarity with or resistance to AI, but because no formal decision had yet been made at the leadership level about how to approach it.

The five stages, briefly, are:

• Stage 1: Curious but Unstructured. Staff are using AI tools, often on personal or free accounts, with little to no organizational visibility, no policy, and no accountability structure.
• Stage 2: Eager Experimentation. Pilots are running in one or two functions, early returns are visible, but the experiments are disconnected from organizational strategy and gains are not scaling.
• Stage 3: Structured Adoption. Leadership is aligned, a formal governance framework is in place, and AI is being integrated into standard operating procedures. Early wins are documented and communicated.
• Stage 4: Workflow Integration. AI is embedded across core workflows, processes are documented and continually optimized, and decision-making is increasingly supported by AI with deliberate human oversight.
• Stage 5: AI-Enabled Organization. Multi-step workflows are powered by AI agents, staff are focused on high-value mission-critical work, and impact is clearly measured and reported.

It is worth being direct about what this model is and is not. It is not a prescription for every organization to reach Stage 5. The right pace and depth of AI adoption looks different depending on organizational size, mission type, staff capacity, and constituent relationships. The goal is not advancement for the sake of advancement. The goal is intentional, mission-aligned progress from wherever an organization stands today.

What the model does make visible, and what the convenings confirmed, is the distinction between AI maturity and AI readiness. Maturity describes where an organization is: the tools in use, the behaviors present, and the results being generated. Readiness describes what it would take to move forward: leadership alignment, a defined strategy, governance infrastructure, and a plan for building skills and accountability across the organization. The two are frequently conflated. The cost of that conflation is an organization that mistakes activity for progress.

The Regional Context

South Florida provides a particularly useful lens for understanding why this moment matters as much as it does.

The region is making serious institutional investments in AI literacy and workforce development. Miami Dade College’s AI Center is expanding free access to AI education for one of the most diverse communities in the country. Vanderbilt University is developing a graduate campus in West Palm Beach with business and engineering graduate degree programs deliberately infused with AI. These sustained institutional commitments, backed with the support of local and state governments, reflect a broader pattern visible across high-growth markets nationally: when a region accelerates its investment in AI infrastructure, academic and philanthropic communities within them respond and scale alongside that investment.

The organizations positioned to earn donor trust and institutional partnership in that environment are not the early adopters, but the ones that can demonstrate operational excellence and measurable impact and the governance structures to sustain both. The competitive advantage is already accumulating for those who are building that foundation now.

The Shift That Matters Most

Among the leaders in these convenings who had moved past the governance question and into genuine strategic planning, a more fundamental reorientation was visible, separating organizations making incremental gains from those building durable organizational capacity.

The default mentality in most nonprofit organizations is thinking of AI as a driver of individual productivity. AI helps a specific employee work faster, draft more easily, or process information they do not have time to handle manually. That value is real. It is also the smallest version of what AI can do for a mission-driven organization.

The more impactful mentality is thinking of AI as a driver of workflow transformation. The relevant question is not how AI makes a specific employee more efficient. The relevant question is this: once an organization has identified where the structural inefficiencies and high-cost, low-leverage work exist across development, programs, finance, HR, and communications, how can those workflows be redesigned with AI embedded from the start rather than added as a layer on top?

Organizations making that shift are not saving one person a few hours a week. They are removing drag from across the organization and redirecting human capacity toward the work that requires human judgment: building relationships, serving constituents, making mission-critical decisions, and driving impact that no tool can produce on its own.

The outcome is not primarily operational efficiency, though that follows. It is mission impact at scale: better program delivery, stronger constituent relationships, more informed leadership decisions, and an organization structurally capable of doing more of the thing it exists to do.

Where to Start

In both convenings, leaders who had arrived expecting a technology and tools discussion found themselves in a change management and governance discussion instead. That reorientation is, in many ways, the most important thing these conversations can achieve.

To reorient leadership’s approach to AI, three actions apply immediately, regardless of where an organization currently sits on the maturity model.

The first is an honest audit. Before any policy can be written or any strategy developed, leadership needs an accurate picture of where AI is already being used across the organization, not just where it has been sanctioned. You cannot govern what you cannot see.

The second is naming an owner. Accountability for AI governance has to be assigned. It does not require a new hire or a new title. It requires a decision about who is responsible and what authority they carry.

The third is drafting a policy. A foundational AI policy is not a compliance document. It is an infrastructure document that needs to answer four questions that most organizations are leaving entirely unresolved:

• Approved Tools: Which AI platforms does the organization officially support, and which are off-limits? Does the organization require enterprise-level tools that include data protection agreements, and have the terms of service for every tool currently in use been reviewed?
• Disclosure: Are employees required to disclose to clients, donors, and colleagues when AI has meaningfully contributed to their work, and is there a clear, shared definition of what meaningful contribution looks like?
• Accountability: What human review process ensures that AI-generated content is checked for accuracy before it is distributed, and who is designated as the organizational champion responsible for AI governance and ongoing integration?
• Training: Who is responsible for training staff, on which tools, and on what timeline? Is that training mandatory for all employees, or targeted to specific roles and departments?

The risk of moving too fast without guardrails is real: confidential donor data runs through unapproved platforms; AI-generated content goes out without human review; staff operate without guidance and without accountability. Organizations that remain ungoverned while the field advances around them will find themselves at a compounding disadvantage.

Questions That Leaders Should Ask Themselves

Before pursuing this framework, leadership should ask itself the following:

1. Do we know which AI tools our staff is using right now, including the ones we did not approve?
2. Has the organization ever had a formal conversation about AI at the leadership or board level?
3. If a donor or constituent asked today whether we use AI, would we have a clear, confident answer?
4. Is there a single person on the leadership team accountable for how AI is being used organizationally?
5. Have we identified even one workflow in the past six months where AI could meaningfully reduce administrative burden or improve outcomes?
6. Do staff have any formal training or guidance on how to use AI responsibly and effectively?
7. If leadership had to place the organization on the five-stage maturity model right now, could they do it honestly, and would the full leadership team agree on the answer?

Three or more “no” or “I do not know” answers are a signal that the organization is ready for this conversation rather than another tool, another demo, or another pilot.

Prioritizing a discussion around these questions at a leadership level is more urgent than learning any AI tool or technology. The leaders who make the decision now to treat AI as a leadership question, not a technology question, will be the ones setting the standard for the rest of the sector in the years ahead.

If you are seeking guidance in navigating your organization’s approach to AI, get in touch with us today.

Lauren Jones Kenny is a Director at Orr Group. Lauren collaborates with nonprofit leaders to understand their needs, develop innovative approaches, and enhance organizational processes to advance missions and accelerate impact.