Is your board ready to govern AI?

For a smallish organisation, it was a significant undertaking. And it was genuinely hard. We underestimated what it would take: the vision for why we were doing it was never clearly communicated, people's roles changed without adequate training, and the assumption that everyone would reach the same level of competency at the same pace turned out to be wildly optimistic. There were also the committed resisters, the people who viewed any digital change as an imposition rather than an opportunity.

I think about that experience often now, as AI moves from novelty to convention across our sector. The technology has changed dramatically. The change management challenge, in many ways, has not.

AI is not just another software upgrade. It raises genuine questions about data privacy, workforce impact, ethics, environmental cost and organisational purpose. Boards cannot afford to leave this conversation entirely to management. Governing AI well is a leadership responsibility, and it starts before a single tool is adopted.

Start with understanding, not enthusiasm

The temptation, especially when AI is everywhere in the news, is to act quickly and figure out the details later. I'd encourage boards to resist that. The first step is not adoption; it is understanding.

Before your organisation commits to any AI tool or strategy, it is worth doing a proper audit of what digital tools are already in use, including tools that staff may be using informally or on their own initiative. From there, a structured conversation about where AI might genuinely add value is far more useful than chasing whatever is generating the most buzz.

That conversation should include a clear-eyed cost-benefit analysis, one that goes beyond dollars and productivity. Consider the ethical implications: who benefits, who might be disadvantaged, and what happens to your people's roles. Consider environmental impact, since AI tools can be energy-intensive, and consider your reputation. Does this align with what your community expects of you?

A SWOT analysis can help structure this thinking. What are your organisation's real strengths and limitations when it comes to adopting AI? What risks exist in moving too fast, and equally, what risks exist in not moving at all?

Know where your people actually stand

One of the lessons I carry from that Office 365 migration is that unaddressed anxiety becomes active resistance. If people feel that change is being done to them rather than with them, the blockers appear. And in the context of AI, the anxieties are often deeper than "I don't know how to use this."

People have legitimate concerns: that their job will be replaced, that they will be left behind, and that the organisation's values will be compromised for the sake of efficiency. These concerns deserve to be taken seriously, not managed away with a reassuring all-staff email.

I'd recommend genuine consultation before any AI implementation. Capture what people are worried about. Understand what support and training they will actually need, not what you assume they need. Set individual or team-level goals so that people can see what this change means for them specifically, not just for the organisation in the abstract.

This is also where the board has a direct role. Directors can model curiosity rather than anxiety. A board that engages thoughtfully with AI, asks hard questions and is visibly committed to doing this well sets a tone that flows through the whole organisation.

Govern the implementation, not just the decision

Once your organisation is ready to move forward, start small and scale deliberately. Piloting one tool in one area of work, reviewing what you learn, and then broadening from there is far less risky than an organisation-wide rollout.

Boards should ensure that appropriate policies are in place before implementation begins, not after. This includes data privacy and security policies, a clear position on what AI tools are and are not approved for use, and guidance on how AI-generated content should be identified or disclosed. For smaller organisations that lack the in-house expertise to develop these, sector resources and templates are increasingly available, including this suite of AI resources from Community Directors.

Regularly review the implementation at both the operational and governance levels. AI tools are not set-and-forget. They change, they are updated, and their implications shift. The board's job is to ensure that management is reviewing these tools actively and that any significant change in use or risk is escalated appropriately.

It is also the board's responsibility to manage expectations. Ambitions need to be calibrated against actual resources and capability. An organisation that commits to an AI-enabled future without the budget, infrastructure or skills to support it is not being bold; it is being reckless.

Build for the long term

Adopting AI is not a project with a finish line. Embedding it well requires ongoing investment in training, regular policy review, and a plan for how new staff will be inducted into your organisation's approach.

Boards should make sure that training budgets reflect this reality. People who were comfortable with a tool 12 months ago may find that the tool has changed substantially. Good practice needs to be championed actively, not assumed.

It is also worth thinking carefully about how your organisation tells this story publicly. AI attracts scrutiny, and not-for-profits are held to a high standard by their communities and funders. If you are using AI in your work, being transparent and thoughtful about how you explain that use will serve you better than staying quiet and hoping no one asks.

The governance bottom line

Be bold, but be strategic. The risk of doing nothing with AI is real, but it is not greater than the risk of doing something poorly.

Purpose must remain central. If an AI tool saves your team time but undermines your mission, your community relationships or your people's dignity, it is not a good investment. The question to keep returning to is not "Can we use AI for this?" but "Should we use AI for this, and does it make us more capable of doing what we exist to do?"

The board's role is not to be the AI experts in the room. It is to ask the right questions, set the right direction and policies, and make sure that the organisation's values are reflected in every technology decision it makes.

That was true when we moved to Office 365. It is even more true now.