Australia's first major report says AI hasn't cut jobs. The real reason is most businesses haven't properly implemented it yet.
On 8 July, the Department of Employment and Workplace Relations released its first major report on how AI is changing Australian jobs. The headline finding was calm. AI is not yet causing mass layoffs. Employment in the occupations most exposed to AI still grew, by 5.6 per cent since late 2022, though slower than the 9.5 per cent in the jobs least exposed to it. Software development roles actually rose by 25 per cent over the same period, and young workers held their ground too.
If you run a business, that reads like good news, and mostly it is. But there’s a plainer reason the layoffs haven’t arrived, and it’s worth sitting with. Most Australian businesses haven’t really put AI to work yet. They’ve bought a chatbot subscription for the team and counted it as done. Used properly, AI doesn’t shrink a team. It lets the same team handle far more. So the report isn’t a sign you can relax. It’s a sign the real work hasn’t started.
Why the layoffs haven’t come
The report was careful with its own findings. It noted that AI-exposed jobs were already growing slowly before ChatGPT arrived, so it stopped short of blaming AI for the gap. It found no rush of people changing careers, which you’d expect if machines were quietly taking over. In its own words, the job market remains strong.
There’s a simple reading of all this. You can’t lose your job to a tool nobody has properly deployed. Handing every staff member a chatbot login feels like adopting AI, but it rarely changes how the work actually runs. A few people use it to tidy an email or draft a note. The rest forget it’s there. So nothing in the numbers moves, up or down.
Take a mid-sized Australian firm, say a [50-person] accounting practice. That figure is only an example to make this concrete, not a real client. Last year they gave everyone a chatbot subscription and told the team to use it. A handful did, here and there. A year on, no one had lost a job, but nothing had grown either. That’s where a lot of businesses sit right now, with no harm done and no ground gained.
What proper implementation looks like
Real AI work looks different from a subscription on every desk. It means building the tools into how the job is actually done, connected to your own systems and your own data, doing real steps of the work while a person stays in charge. This is the sort of thing AI agents are built for. Not a chatbot people visit, but quiet systems that handle the repetitive parts of a workflow from start to finish.
Go back to that accounting practice. Instead of a login nobody opens, picture agents that read incoming client documents, sort them, pull the figures into the right place, and flag anything unusual for a person to check. This is the kind of repetitive work that AI agents for professional services are suited to. The work that used to eat a junior’s whole week gets done in the background. Nobody was let go. The same team simply took on more clients, the kind of growth the firm had been turning away for lack of hours. That’s what proper use delivers, and it’s why real implementation grows a business rather than cutting it.
A fair word of caution here. AI won’t fix a mess for you. Point it at disorganised records and it will process the disorganisation faster, not clean it up. It doesn’t replace judgement either. When client data or private information is involved, a person stays accountable, and your obligations under the Privacy Act 1988 don’t change because a machine did the first pass. The gains are real, but they come from careful work, not a switch you flip.
The Australian reality
The report put Australia in an unusual spot. The United States is seeing its job market weaken, while ours has held firm. That gap is a window, and windows close.
Toby Walsh, an AI professor at UNSW Sydney, made the point plainly in response to the report. “When the numbers are showing significant impact of AI on jobs, it’s already too late to respond,” he said. “We actually have to be proactive.” He was talking about governments preparing workers, but the same logic runs through any business. By the time AI shows up in your competitors’ results, the lead is already theirs. The firms that built the infrastructure early will be the ones taking on work everyone else can’t reach.
What this means for you
The government’s report is reassuring, but it’s easy to take the wrong lesson from it. The absence of layoffs isn’t proof that Australian businesses have handled AI well. For most, it’s proof they’ve barely begun. That’s not a reason to panic, and it’s certainly not a reason to cut staff. It’s a reason to start the real work while the market still gives you room to.
The gap between a chatbot subscription and a system built into your business is the whole game, and it’s where the growth lives. If you’d like to find where that growth sits for your team, take our AI Roadmap Interview. You’ll talk through your goals, your workload, and where your people lose the most time, then get a custom plan for where to start.
