AI at Work: Don’t Just Automate the Job — Redesign It
Written by Cherilyn Walthew
AI is changing how work gets done. But for most New Zealand employers, the useful question isn’t simply whether AI will replace jobs.
It is:
Which parts of the job should AI do, and which parts are better done by people?
That shift in thinking matters. Deloitte New Zealand’s 2026 Human Capital Trends research found 87% of New Zealand organisations consider planning for future workforce needs amid disruption very or extremely important. Yet only 2% say they are leading when it comes to intentionally designing how people and AI work together.
In other words, employers know change is coming. Working out what that actually means for their people is proving harder.
Start with the work, not the technology
Instead of starting with: “Where can we use AI?”
Try: “What are our people spending time on that doesn’t really need a person?”
That might include repetitive administration, first drafts, summarising information, basic reporting or pulling together research.
If AI can take some of that workload away, great. But then ask what your employee could be doing with the time you have freed up.
A manager who spends less time preparing reports could spend more time coaching their team, solving problems, talking to customers or improving how work gets done. That is where AI can start adding real value.
Jobs are changing — and so are the skills we need
PwC New Zealand’s 2026 AI Jobs Barometer suggests the bigger impact of AI may be job redesign rather than job replacement.
It found that occupations highly exposed to AI have added an average of 78 new skills per role since 2019. As AI handles more administrative and repetitive work, PwC expects human work to increasingly focus on problem solving, judgement, creativity, collaboration and relationship management.
Those are the skills businesses have often referred to as “soft skills”. There is nothing particularly soft about them. Good judgement, communication, adaptability and the ability to manage people are often what separates an average employee or manager from a very good one. AI may produce an answer in seconds. A person still needs to know whether that answer makes sense.
Don’t accidentally deskill your team
There is a flip side. If AI starts doing too much of the drafting, analysing and problem-solving, employees can lose opportunities to develop those skills themselves. This is particularly important for junior staff. Think about the jobs people traditionally learned by doing: preparing the first draft, researching the issue, working through a problem or watching an experienced employee make decisions.
If AI now does all of that, where does the learning happen? Employers need to think about what skills people still need to develop, where human judgement must remain and who is responsible for checking AI-generated work.
SMEs have a particular challenge
Workforce 2030, powered by the Employers and Manufacturers Association, reports that smaller New Zealand businesses are actively experimenting with AI but are less likely to have formal governance or workforce capability programmes in place. It also identifies increasing demand for judgement, adaptability, communication, ethical reasoning and critical thinking as routine tasks become automated. That feels particularly relevant for SMEs. You probably don’t need an “AI transformation strategy” and a 40-page implementation plan. You do need to know what your staff are using, where it adds value, where the risks sit and what you expect people to remain responsible for.
Your job descriptions may need to catch up
If AI materially changes how a job operates, eventually the paperwork should reflect reality.
Ask:
Which tasks have reduced or disappeared?
What new responsibilities have emerged?
What skills are becoming more important?
Has decision-making or accountability changed?
Are you still measuring the right things?
This doesn’t mean rewriting a job description because someone starts using ChatGPT. It means recognising when the job itself has genuinely changed.
Where EASI NZ sits
For most businesses, the opportunity is fairly simple: Use technology to remove low-value work, then make better use of the people you already have. That could mean redesigning roles, updating job descriptions and KPIs, developing managers or putting more effort into the human skills technology cannot easily replicate. AI can make work faster. Good job design can make the business — and the people in it — better.
If AI is starting to change how work gets done in your business, EASI NZ can help you work through what that means for your roles, responsibilities and people before the technology starts driving the structure.
