A Mercer study reported this week by News.com.au found that six in ten Australian companies expect to lose up to one in five jobs within two years because of AI. That number is getting attention. What is not getting enough attention is what it actually tells us about how Australian businesses are approaching this transition, and what it means for the people already sitting in those roles right now.
The Finding and What It Really Says
The Mercer data, as reported by News.com.au, points to a significant share of the Australian business community anticipating meaningful workforce reductions tied to AI adoption within a short timeframe. Two years is not a distant planning horizon. It is the next budget cycle, the next performance review season, the next round of team restructures.
Before we talk strategy, it is worth pausing on what that number actually represents in human terms. If you run a team of twenty people and you are in that 60%, you may be looking at up to four roles changing or disappearing. Those are not headcount figures on a spreadsheet. They are people with mortgages, with families, with years of institutional knowledge in your business. They deserve more than being the last to know.
That framing matters because the data is being reported largely as an economic and productivity story. It is also a people story, and businesses that treat it as only the former are already setting themselves up for problems the technology will not solve.
Job Losses from AI Are Not Inevitable, They Are a Default Outcome
Here is the thing that gets buried in these headlines. Workforce reduction is not a guaranteed output of AI adoption. It is what happens when businesses automate first and plan for their people second.
When a business identifies a task that AI can handle and simply removes the person doing that task, the outcome is predictable. But that is a choice, not a law of physics. The businesses getting better outcomes right now are the ones that approach AI adoption differently. They are asking which parts of a role can be augmented, not which roles can be eliminated. They are identifying where automation frees up human capacity for higher value work. They are treating the transition as a redeployment and upskilling challenge, not a headcount reduction exercise.
That is not idealism. It is practical workforce planning. And it tends to produce better AI adoption outcomes as well, because teams that are brought into the process rather than blindsided by it are far more likely to actually use the tools well.
The 70% of AI adoption failures that trace back to human and process challenges rather than technical ones do not come from bad technology. They come from teams that were not prepared, not consulted, and not given a clear picture of what was coming for them. Fear, resistance, and disengagement are entirely rational responses when people feel like decisions are being made about their working lives without them.
Australia Has a Workforce Transition Planning Problem
Australia does not have a shortage of AI tools. It does not have a shortage of vendors, webinars, or LinkedIn posts about productivity gains. What it has a shortage of is businesses treating the workforce transition with the same seriousness as the technology rollout.
The pattern across industries is consistent. Businesses move quickly on the technology decision, sometimes because a competitor did, sometimes because a vendor made a compelling pitch, sometimes because a senior leader came back from a conference energised. The procurement and implementation get resourced. The people planning does not.
Workforce transition planning is not a soft, optional add on to an AI implementation. It is the part that determines whether the implementation actually works. It covers things like which roles are changing and in what ways, how affected team members are being communicated with and when, what retraining or redeployment pathways exist, how the business will measure success in human terms not just efficiency terms, and what support structures exist for people navigating genuine uncertainty about their futures.
None of that is complicated in concept. Most of it is just applying the same rigour to people that businesses already apply to systems. But it requires a deliberate decision to start with people rather than treating them as something to manage once the technology is already in place.
What the Businesses Getting This Right Are Doing Differently
The businesses that are navigating AI adoption well right now share a few consistent characteristics. None of them are about having better tools or bigger budgets.
They start the people conversation before the technology conversation. Before a single vendor is evaluated or a licence is purchased, they have mapped which roles will be affected, what those changes will look like, and how they will communicate that to their teams. Their people hear it from leadership before they hear it through the grapevine.
They separate the automation decision from the headcount decision. Not every task that can be automated should result in a role being removed. Many businesses are finding that AI frees up capacity that, when redirected well, makes existing team members significantly more valuable. The question worth asking first is not which roles can we reduce, but which capabilities in our existing team can we amplify.
They treat adoption as an ongoing process, not a project with an end date. AI tools are changing quickly. The way a team interacts with those tools twelve months after implementation looks very different from the way they interacted with them in week one. The businesses getting sustained value are the ones that build continuous learning and adaptation into how they work, not just into the initial rollout.
They measure the human outcomes alongside the efficiency outcomes. Adoption rates, team confidence, role satisfaction, and retention are as important as time saved or cost reduced. A business that automates a function and loses three experienced team members to resignation in the following six months has not won on the numbers, regardless of what the productivity data shows.
If Your Business Is in That 60%
The question worth sitting with right now is not which roles in your business might change because of AI. That conversation is already happening in most leadership teams. The more important question is whether your people know what comes next for them.
If the answer is no, or even not yet, that is the thing to address first. Not the tools. Not the efficiency case. The people who will be most affected deserve to be brought into this conversation before the decisions are made, not after.
The window to do this well is not closed. The businesses starting that conversation now are not behind. They are making the choice that separates the companies that will look back on this period as something they handled well from the ones that will look back on it as something that happened to them and to their teams.
In five years, the companies that won will not be the ones that adopted AI first. They will be the ones that brought their people with them. The Mercer data is a prompt to start treating that as the real priority, not a trailing concern once the technology is already live.
Get in touch at www.xsiv.au/#form