top of page
Search

What's different about change in an AI world?

What an interesting start to the new financial year we have had in the world of AI.


From the big tech AI battle continuing with new models being dropped within hours of each other, to Prime Minister announcements on the future of AI for Australia.


Here are two other pieces of news that landed this month that, side by side, which also tell you almost everything about where AI has actually got to.


The first: Microsoft committed $2.5 billion and around 6,000 of its own experts to a brand-new business, Frontier Company, built to sit inside its customers and help them genuinely put AI to work. To handle the integration, the configuration and, in Microsoft's own words, the change management. The biggest software company on earth has decided the hard part of AI is no longer the model. It is the change around it.


The second, closer to home: Deloitte's State of AI 2026 survey found that while around 69% of Australian organisations are already using AI agents, only 22% have an advanced way of governing them. The same gap shows up worldwide: 85% of companies plan to tailor AI agents to their business, and just 21% have a mature model for governing them.


Technology capability is racing ahead. Money is racing ahead. Readiness, the human and governance side, is the gap. And that is the side we can actually do something about.


This is exactly the gap Nuovo Insights works with leaders to address. This week it is exactly what we addressed with a room of HR Directors in Melbourne.


Why the old change playbook needs a rewrite


Every wave of technology most of us have led before changed the tools. AI is different. AI changes the work itself. When software can pursue a goal, make a decision and take an action on its own, the question stops being "how do we train people on the new system?" and becomes "what is a human role now?"


The well-known and used change models are still brilliant models, and change discipline still matters. Prosci's research shows initiatives with excellent change management are seven times more likely to meet their objectives than those with poor change management. The method is not the problem. The assumptions underneath it are.


Four things make AI change genuinely different.


  1. Change does not start at the top. For once, the change did not begin in the boardroom. Your people are already using AI, often ahead of any plan. In Australia, 66% of workers already use at least one AI tool regularly, and more than half of them reach for tools their employer has not formally approved. Leadership's new job is to lead something the workforce started.


  2. Identity first. Old change asked people to adopt a new process, technology or policy. AI asks something deeper: if the thing I am known for can be done by a machine so quickly, who am I now? There is a paradox of people are using AI of their own choosing, but when being done by the organisation they disengage and do not trust it, as they cannot see themselves in the new picture. They protect what they have. Belonging, status and trust are exactly the things technology cannot hand you, and exactly where good leaders earn their keep.


  3. The target keeps moving. The old model had neat phases and a finish line. You cannot "finish" adopting AI. That calls for planning in shorter loops, quarters and sprints, and for helping people feel steady on ground that keeps shifting under them.


  4. Fatigue is real, and implementation pace is a choice. Gallup put global employee engagement at 21% in 2024, the lowest since 2020, with managers disengaging the most. Those are the very people we ask to carry the change. You can plan change to protect your people, or to burn them out.


The Nuovo Change Model: keep what works, rewire the assumptions


At Nuovo Insights, drawing on our psychology and organisational change background, we redefined the loop. We call it the Nuovo Change Model, addressing the four domains of business strategy, people strategy, operating model and governance, and it runs as four moves around a human core of judgement, relationships and expertise.


- Sense: find where AI is already moving in your organisation, and where the real value sits.

- Shape: co-design roles, guardrails and ways of working with people, never to them. People commit to what they help create.

- Embed: make the new way normal inside daily work, not a one-off launch.

- Adapt: learn, adjust and go around again as the technology moves.


It never refreezes.


Where to start: the first ninety days


You do not need Microsoft's $2.5 billion or its 6,000 experts. You need the same discipline at your own scale, and the first 90 days are doable.


Month one, sense: map where AI is already being used, and listen for the identity questions underneath. Month two, shape: co-design guardrails your people can actually follow, and redesign one role with one team. Month three, embed: make the new rhythm normal, share honestly what worked, and set the next loop in motion. Small, human and continuous is better than big, technical and one-off, every time.


If you run a smaller business, the same logic shrinks to fit. You do not need a governance model or a $2.5 billion budget. You need one honest habit, started this week: a human checking the work, and being open about where AI helped.


Where to?


AI is not a replacement for your judgement, your relationships or your expertise. Lead the change as though that is true, because it is. That, to me, is what making AI human looks like.


So here is the first question, and you can answer it this week: where is AI already moving in your team, with or without a plan? Our free AI Readiness Assessment is a good place to start.


*Tracy Attanasio is Founding Principal of Nuovo Insights, an Australian AI change advisory firm. She has spent 25 years leading large scale change, and holds degrees in psychology because AI adoption is a people question first.*


Sources

 
 
 

Comments

© 2025 Nuovo Insights

  • Instagram
  • LinkedIn

ABN: 75690323418

bottom of page