By: Darryl Lee and Ashley Pettifer, Ryelle AI Portfolio Leads 

Most AI governance advice starts in the same place: draft a policy, build a risk matrix, and decide which tools are approved. It sounds responsible, and for a bank or a law firm, it probably is. But for nonprofits and mission-driven organizations, that order is backwards. Starting with compliance can put the focus in the wrong place, quietly importing priorities and assumptions that may not reflect your mission, your values, or the communities you serve.

Here’s the catch: a risk matrix is designed to protect the organization. Your mission is designed to serve people. Those aren’t always the same thing.

Take an AI tool that summarizes client conversations. It can save hours and even flag issues staff might miss. But it can also miss the context behind what someone is saying. For most businesses, that’s a small tradeoff. For nonprofits and public sector organizations, that context is often what matters most.

So don’t start with the technology. Start with people.

Ask who is in the room

AI decisions in most organizations are made by a small group, usually leadership plus whoever is closest to the budget. IT and HR are often left out, or brought in late, because they get treated as enabling functions rather than partners in strategy. Frontline staff are rarely there at all. That is where governance starts to go wrong, and it happens before a single tool is chosen.

The people you leave out are the ones who see what the tool will actually do. IT understands what it touches and what it puts at risk. HR understands what it does to roles, workload, and the shape of a job. Frontline staff understand the work the tool claims to speed up, and they are the first to notice when a faster answer is a worse one. Leaving any of them out does not simplify the decision. It just means you make it with less information.

A better group is leadership, IT, HR, and the people who use the tool every day, with someone in the room who can speak for the community you serve. You do not need a huge committee. You need the absence of those voices to feel strange rather than normal.

Ask what it does to the relationship

Mission-driven work is relational in a way most commercial work is not. Trust is the product and the mission. So the governing question is not just whether a tool is accurate or secure, it is what the tool does to the relationship between your organization and the people you serve.

An AI tool can finish a task faster while making a relationship thinner. It can (sometimes) answer a question correctly while still making someone feel like they weren’t really heard. Speed and trust are often traded against each other, and mission-driven organizations are the ones who cannot afford to trade off on trust. Name the cost to trust out loud, every time. Don’t let speed win by default just because it’s easier to put on a slide.

Make the quiet costs visible

The benefits of AI arrive immediately: hours saved, backlogs cleared. The costs arrive slowly and off the books. Staff who stop developing judgment because the tool offers an answer first. Knowledge that used to live in people now living in a system you do not control. A subtle shift in who gets to decide what counts as a good summary, a strong application, a real need.

Good governance is really just the discipline of putting the slow costs next to the fast benefits, and weighing both at the same time. Sometimes the tool still wins. But it should have to earn the win.

Govern where people can see it

Compliance is something you do for a regulator. Governance, done well, is for the public. Tell the people you serve when AI touches their case. Tell your staff how these decisions get made and how they can challenge them. A mission-driven organization that hides its AI use is borrowing a posture that does not fit and will not hold.

None of this requires a data science team or a six-figure platform. It requires putting people ahead of the policy document, and being willing to turn off a tool (even if it works) when it pushes you away from who you are.

The organizations that get this right will not be the ones with the longest policies. They will be the ones that can still tell the difference between a summary that is accurate and a summary that is true, and that trust their people to push back when there’s a gap.

Make that instinct deliberate. Share it across staff. That’s the framework.

 — 

Ryelle Strategy Group works with public sector organizations, corporations, and NGOs to navigate strategy, transformation, and organizational change. If these questions are showing up in your organization, we’re happy to continue the conversation: www.ryellegroup.com.