Leading people through AI change

On Monday morning the leadership team meet. Customer demand is up, board expectations are higher, and everyone’s running into the same tension; AI is already in the building. It’s in pilots, vendor roadmaps, and your people’s day-to-day, yet the way work flows hasn’t really changed. Your team feels the pull to go faster and the weight of getting it right. That’s the moment of time we’re in.

This isn’t a tools rollout or a Cloud transformation. It’s a fundamental shift in how we deliver value to customers and communities, and how we look after the people doing the work. This places new challenges on leaders and executive teams, because how do you lead your people through fundamental change?

The goal isn’t to make yesterday’s complex processes run faster, it’s to change how we work, using the full suite of data to recommend the next best step and take action so every customer interaction is simple, clutter-free, and effective.

Start with the people, stay with the people

AI is often met with mixed emotions - excitement, curiosity, scepticism, fear. If we want people to try, learn and adapt, they need psychological safety and a story they can believe.

Offer a simple change narrative that you can repeat without the slides:

• Why now? (Outcomes for customers and whānau)
• What won’t change? (Our values, Privacy Act 2020 obligations, data sovereignty)
• Who will be impacted? (Change impacts, what upskilling is required, workforce planning)
• What will change? (How we work, how decisions are made, how we measure value, how we learn).

Consistent leadership and messaging from the top steadies the ship more than any framework or PowerPoint pack.

Make changes and learning visible and legitimate and stay engaged in the feedback that underpins effective AI. Notice outcomes, test, refine and repeat. Celebrate good questions, prompts and catches because when leaders model curiosity, it becomes contagious.

For example; each morning, a public sector Deputy Secretary uses AI to scan and summarise her calendar, key policy updates, and interagency developments. The AI flags strategic risks and opportunities based on recent decisions and system signals. She uses this to shape her priorities for the day and shares a short reflection with her senior leadership team: “Here’s what I’m focusing on today, and one thing I’m testing with AI - let me know what I’ve missed.”

One team, one customer

The business–technology separation is a dying concept. Customers don’t care which side of the org chart fixed their problem; they care that it was solved quickly, fairly, and once. When considering the introduction of new ways of working with the team, work collaboratively with your technologies partners to really lean into ‘art of the possible’ thinking. Focus product owners, policy minds, designers, engineers, HR and transformation teams and operations on the same outcome and ask: What’s the fastest and safest path to delivering value for our customers? When that group shares one backlog, one definition of “done”, and one set of measures, the conversation changes from “handover” to “together”.

Lead uncertainty by making trade-offs visible

Much team anxiety comes from silent trade-offs. Say them out loud. We’re choosing to move faster here because the risk is low and reversible. We’re choosing to automate rote tasks to free people up for higher value work. We’re choosing to move slower here because the impact is higher and we need to bring communities with us. That’s leadership; naming the tension, then owning the pace.

Be transparent in plain language about what AI is used for, the data involved, how this will impact people and teams, and how to challenge a decision. That’s how you keep your social licence.

Measure change - performance and value - beyond speed

Yes, cycle time and cost-to-serve matter. So do right-first-time, equity of outcomes for priority segments, energy use per transaction, and customer confidence. When you broaden the definition of value, teams make better choices. For example, tie wins to FinOps and sustainability goals so your CFO and your climate commitments pull in the same direction.

Look after the people doing the mahi (work)

Roles will evolve. Managing uncertainty and resilience to change will be critical for individual success through the transition. To support the evolution and uncertainty, recognise new responsibilities (prompting, validation, model stewardship), and celebrate wins. Invest in capability uplift, not just technical skills, but also change resilience, ethical awareness, and collaborative ways of working. Create psychological safety for experimentation and learning. See the ‘Hot Takes’ box where we have some tips for supporting teams through the transition.

Making it real – a short example

Think about a customer querying a decision - this may require information from multiple teams to synthesise data and respond.

Today, a case might bounce across three teams, with good intentions and slow outcomes. Tomorrow, an AI assistant could pull together information from across knowledge sources, summarise the output and flag edge cases, or inconsistencies in the original decision, not to replace the case manager, but to put their judgement where it matters most.

Customers get faster updates and an easy way to request review. Case managers spend time on complexity and care. Leaders see fewer late responses, stable complaints, lower unit cost, and less burnout. Same mission, but with a better flow.

Team conversations to move the needle

In your next leadership forum with your team, try these:

• What will “good” look like for our customers, beyond speed? Where do our humans add value for our customers that AI cannot?
• Where does human judgement change outcomes, and where are we adding unnecessary AI-human hand-offs?
• How will we protect privacy, consent and kaitiakitanga as we roll out AI?
• What’s our plan to help people learn in the flow of work with time, pairing and support?
• How will we know if our people have trust in our AI transition?

Final thought

AI will keep shifting under our feet. Your job as a leader isn’t to predict every turn, but to hold the centre - the customers you serve, the people you lead, and the values you won’t trade. That means bringing business and technology together as one team, using ‘art of the possible’ thinking to reimagine how the work evolves. Keep humans in the loop where judgement truly changes outcomes, be open about the trade-offs you’re making, and invest in learning you can see and feel on the floor. Do that, and your organisation won’t just absorb AI, it will become better at serving people in Aotearoa, together.