How do you bring AI into an organisation whose whole purpose is human?

Client: A national children's charity
Engagement: Technology Capability and Future Readiness, April to August 2026
Role: People & Culture and AI consultant

The question

In April 2026 the CEO asked three practical questions.

How do our teams use technology today?
What capability will our people need for what's coming?
Where can systems, automation and AI give time back to the mission?

The key question was how an organisation built on human connection takes up AI without losing what makes it work.

How I approached it

To me it was a question of culture before being a question of technology. The project ran over four months, with a few focus groups and listening sessions each week. I listened to teams across the whole organisation, from the Leadership Team to the frontline, and asked people to talk me through how their work actually happens: where the time goes, what they've tried, and what they'd change.

Every team's story was created and understood before the whole picture, and the plan built on it, went to the CEO and the Leadership Team. I worked in partnership with AI throughout, and I checked everything before it went anywhere. That's the same principle I went on to propose for the charity.

What I found

The teams were curious and engaged. Almost every team could name where AI or a better system would help, and some were already well ahead: building their own tools, using AI as a critic of their drafts, or drawing on it daily for analysis. Others were exploring for the first time. Readiness sat on a continuum, from concerned through cautious and occasional to regular and embracing.

Some team members said they wanted to learn this for themselves, and that they didn't want to lean on AI blindly. This indicated that a system that builds capability first is more effective than just enforcing a policy.

The opportunity is large. In every team, time was going into re-keying, chasing, searching and rebuilding: work that AI assistants can carry once the data underneath is right. Every hour returned is an hour with a child and family, a donor or a volunteer.

One team gave the work its best principle. In their words: "The whole purpose of [our work] is to capture the child's likes, interests, their joy and excitement. And if we were to use AI to correct that stuff, we actually lose the value of what we're trying to achieve." So the child's voice is never machine-corrected.

What I proposed

The destination is people and AI assistants working in partnership, with people firmly at the helm. Assistants take the repetitive, structured work. People keep the judgement in the work with each child, the care in a family conversation and the relationship with a donor. I offered the Leadership Team one picture for it: an AI assistant is a new kind of team member who never sleeps, and like any new starter it needs a clear role, an induction, supervision and a person accountable for its work.

Technology on its own doesn't change an organisation. The plan treats the charity as a system made up of levers that reinforce a culture of adoption:

leadership accountability
data and information foundations
the people processes, from recruitment to recognition
everyday ways of working.

Each one reinforces the others. Most of the work is people work, and people are what this charity does best.

From the team's input, two aspects have been developed. Eight principles for how AI is used, starting with the mission comes first and a person always checks what an assistant produces. And a 90-day plan that begins with the Leadership Team, then a champion in every team, then the first pilots built from employee input.

Where it stands

The report went to the CEO in August. In September I presented the employee feedback, and the direction the CEO and the Leadership Team had endorsed, to all staff. Copilot is being made available to every staff member inside the Microsoft 365 the organisation already has, and an AI huddle is forming, focused on speed, effectiveness and alignment.

Four measures were agreed up front: hours returned to mission work, data quality, movement along the readiness continuum, and staff confidence.

The design goal throughout was capability the organisation keeps: decisions owned at the Leadership Team table, champions in every team, and no dependence on the consultant who started it.

For your own organisation

If your people are curious about AI and unsure where to start, listening to them is a good place to begin. In my experience the people already ahead of you are the best guide to what's possible. Where are yours? I'd be glad to talk.