How small business and government teams can adopt AI with confidence

It's easy to feel overwhelmed by the pace of AI. Every week, there's a new tool, a new model, and a new headline promising to revolutionise work. But for small to medium organisations and small Government agencies, the path to AI adoption looks a little different. You're not sitting on a massive innovation budget, and you likely don't have a team of data scientists waiting in the wings.

And yet, the opportunity is just as big, if not bigger.

Recently, I listened to a brilliant expert panel exploring what it takes to adopt AI in real-world organisations. A big shout-out to Caelan Huntress for facilitating this discussion as part of Tech Week.

The insights from the panel were clear: you don't need to chase the hype. What you do need is a people-centred, practical approach that works with your culture, your constraints, and your purpose.

So here's what I learned from the discussion, and what I want you to know, about how you can adopt AI with confidence.

AI adoption is a change programme, not a tech rollout

AI isn't a tool you plug in and walk away from. It changes how people work. It introduces new habits, new ways of thinking, and (at least initially) a good dose of uncertainty.

That's why adoption isn't just a technical exercise. It's a shift in mindset. It requires curiosity, experimentation, and above all, support. Leaders who treat it like a traditional software rollout "we'll switch it on and you'll start using it" often run into resistance or apathy. Not everyone hates change. But most people don't feel safe to try something new.

The most successful teams make AI adoption feel safe, useful, and human.

A people-first approach is the real secret

There's no one-size-fits-all AI playbook, but here's what works, especially for smaller organisations and public sector teams.

Start with real problems, not shiny tools

Pick two or three pain points that people genuinely care about. Look for repetitive, frustrating tasks where AI can save time or improve clarity, like summarising meetings, finding information buried in documents, or generating first drafts.

Build trust through transparency

Choose tools that are grounded in your data and show where their answers come from. RAG (retrieval augmented generation) tools can link back to your source documents, helping people feel confident that the AI's output is reliable, not made up.

Create a space to experiment

Adoption is faster when people feel like it's okay to try, to fail, and to learn. This means:

• letting people explore without fear of being judged

• giving them permission (explicitly!) to use the tools

• and carving out a little time to play, even 20 minutes a week makes a difference

Involve people at all levels

Run small pilots with a mix of roles: a team leader, someone from the frontline, someone in comms or support. This builds buy-in, spreads knowledge, and stops AI feeling like something that's being "done to" people.

What's holding teams or organisations back?

Most of the fear or hesitancy we see falls into three buckets:

• "We don't have the tools or budget". Start small. Use free tools like ChatGPT (with privacy in mind) or trial no-code options built for small teams.

• "We're too busy for this". If you don't make space, it won't happen. Transformation can't sit on top of an already overloaded day. Create a little room for people to experiment.

• "People don't trust it". That's normal, especially in Aotearoa New Zealand. Trust is low, and people are cautious. The way to build trust is through experience. Let them see the output based on their own words. Let them test it, question it, and validate it for themselves.

What can leaders do right now?

If you're a people leader in a small to medium organisation or public sector team, here are 5 things you can do this month to get started:

• Pick one team and run a 4-week AI pilot - choose a real task (like meeting notes or research summaries).

• Identify your internal champions - people already curious about AI who can support others.

• Normalise safe-to-fail learning - say out loud "You're allowed to test this, and it's okay if it doesn't work the first time."

• Check your privacy settings - especially if you're using tools like ChatGPT. Be mindful about what data you share.

• Focus on impact, not hype - what saves time, improves service, or removes friction? That's your north star.

Final thought

You don't need to be a tech giant to harness AI. In fact, the most exciting progress is often happening in places like yours - schools, local councils, healthcare providers, frontline service teams. What makes the difference isn't the budget. It's the mindset.

Curious, Cautious. And committed to people first.

If you want to chat about where to begin, I'd love to hear from you.

Let's make AI useful, not overwhelming.