Over the past year, AI has become one of the most frequent topics of conversation among nonprofit leaders, with discussions quickly graduating from what AI is to how to use it effectively. A year ago, many organizations were focused on simply learning about the tool. The questions today are far more practical:

How do we use it responsibly? Where should we start? What are other nonprofit organizations doing?

I have been fortunate to explore these questions from several perspectives, through conversations with clients, moderating discussions with nonprofit finance professionals and listening to technology leaders who are thinking deeply about where AI is headed. Although every organization is at a different stage of its AI journey, I have been struck by how often the same themes emerge.

One presentation that reinforced many of these observations was delivered by Chief Information Officer Robert Stein of the National Gallery of Art. His perspective aligned remarkably well with what I have been hearing among nonprofit executives: AI is more than a new efficiency saver or glamorous technology to implement. It represents an opportunity to rethink how organizations work.

Here are five lessons I believe nonprofit leaders should keep in mind.

LESSON 1: DON'T WAIT FOR THE PERFECT AI STRATEGY

Many leaders feel they need a comprehensive AI strategy before allowing staff to experiment. In reality, the organizations making the greatest progress are learning by doing. They begin with small, low-risk pilots that help employees understand both the strengths and limitations of AI.

Whether it is drafting a first version of board materials, summarizing meeting notes, organizing research or assisting with financial analysis, small experiments build confidence while revealing where AI adds value and where it does not.

Robert Stein emphasized the importance of encouraging experimentation, and I could not agree more. The goal is not to automate everything overnight. It is to help people become more comfortable using AI thoughtfully.

Practical takeaway: Choose one repetitive task your team performs every month. Test AI in that workflow, keep human review in place and document what you learn before expanding further.

LESSON 2: AI LITERACY MATTERS MORE THAN CHOOSING THE "BEST" TOOL

One thing I have learned is that nonprofit leaders spend far too much energy worrying about which AI platform will "win." Given how quickly this technology is evolving, that is probably the wrong question to ask.

The organizations making the most progress are investing in AI literacy rather than becoming attached to a particular platform. They are helping employees learn how to write better prompts, evaluate AI-generated output critically, understand the limitations of the technology and recognize when human judgment must prevail.

During a recent roundtable with nonprofit CFOs, one participant compared today's AI landscape to a horse race. Today's leader may not be tomorrow's leader. That observation resonated with the entire group because it shifted the focus away from the technology itself and toward building adaptable teams.

Practical takeaway: Encourage employees to share prompts, successful use cases and lessons learned. Building organizational knowledge will outlast any single AI tool.

LESSON 3: GOVERNANCE ENABLES INNOVATION

Perhaps the greatest concern I hear from nonprofit leaders is regarding the risks around AI. Questions about confidential information, donor data, internal controls and governance are entirely appropriate. But I have come to believe that governance should not be viewed as something that slows innovation. It should create the confidence to innovate responsibly.

One theme surfaced repeatedly in both my client conversations and a recent CFO roundtable: governance and experimentation should grow together. Organizations do not need a perfect AI policy before beginning to explore low-risk use cases. They do, however, need clear expectations around approved tools, data privacy, human oversight and accountability.

As an auditor, this feels familiar. Strong governance does not eliminate risk; it helps organizations manage it thoughtfully.

Practical takeaway: Develop a simple AI Safe Use Policy before AI adoption becomes widespread. Start with practical guidance, then refine it as your organization's experience grows.

LESSON 4: RETHINK WORKFLOWS, NOT JUST INDIVIDUAL TASKS

Much of the conversation around AI focuses on productivity. “Can AI write an email faster? Summarize a report? Draft a policy?”

Those are useful applications, but I think the bigger opportunity lies elsewhere. Several nonprofit leaders I have spoken with have discovered that AI delivers the greatest value when it improves an entire workflow rather than a single task. 

Preparing board materials, responding to audit requests, drafting grant reports and analyzing financial information all involve multiple steps. Improving the overall process often creates much greater value than simply accelerating one piece of it.

One observation from our roundtable has stayed with me: start with the business problem, not the technology. Instead of asking, "Where can we use AI?" ask, "What is slowing us down?" That simple shift often leads to much better solutions.

Practical takeaway: Identify one process that consistently frustrates your team. Then ask whether AI can reduce friction across the entire workflow — not just one task within it.

LESSON 5: CURIOSITY MAY BE YOUR GREATEST COMPETITIVE ADVANTAGE

For nonprofit organizations to make progress on AI, they do not necessarily need the largest budgets or dedicated technology teams. They just need to create cultures where people are encouraged to learn.

Nearly every organization has employees who are naturally curious and willing to experiment. Those individuals often become informal champions who help colleagues build confidence, share successful approaches and avoid common mistakes.

Leadership plays an important role here. When leaders demonstrate curiosity rather than fear, they give others permission to do the same. AI will continue to evolve. So will the opportunities and risks. Organizations that build adaptable cultures will be far better positioned than those waiting for certainty before taking the first step.

Practical takeaway: Identify one or two AI champions within your organization and create opportunities for them to share what they are learning with others.

QUESTIONS EVERY LEADERSHIP TEAM SHOULD BE ASKING

Rather than trying to predict where AI will be in five years, I encourage nonprofit leaders to focus on questions they can answer today:

• What business problem are we trying to solve?
• Are our data and processes strong enough to support AI effectively?
• What information should never be entered into an AI tool?
• Do employees understand when human review is required?
• What governance and oversight should our board expect?
• How are we helping staff build AI literacy?

These conversations often provide a clearer starting point than discussing technology alone.

LOOKING AHEAD

If there is one thing I've learned from clients, peer discussions and industry leaders over the past year, it is this: AI is not replacing leadership, it is raising the importance of leadership.

Technology will continue to evolve faster than any of us can predict. The organizations that benefit most will not necessarily be those with the newest tools or the largest technology budgets. They will be the ones that stay curious, build strong foundations, establish thoughtful guardrails and remain willing to learn.

That is encouraging because those qualities have always defined strong nonprofit organizations. AI simply gives us another opportunity to put them into practice.

This article was inspired in part by a presentation by Robert Stein, Chief Information Officer of the National Gallery of Art, and by ongoing conversations with nonprofit leaders who are navigating the opportunities and challenges of AI adoption.