The Great AI U-Turn
Three months into 2026, Australian companies have already cut more than 4,450 tech and corporate jobs — with every major round citing AI as the driver. WiseTech slashed 2,000 roles. Atlassian cut 1,600. CBA trimmed 300 while launching a $90 million "Future Workforce Program" in the same breath. Telstra dropped 650. Optus cut another 200-300.
Sydney ranked third globally for tech layoffs, behind only San Francisco and Seattle. AI was cited as the primary factor behind every one of these announcements.
But something the headlines aren't covering: most of these companies are already bringing people back.
Two in Three Are Bringing Workers Back
A February 2026 survey of 600 HR professionals who conducted AI-related layoffs tells a very different story from the press releases. Of the organisations that cut staff for AI reasons, 68.3% have already started rehiring for the same roles they eliminated.
The details are striking:
- 35.6% had rehired more than half the positions they cut
- 32.7% had rehired between a quarter and half
- 52.1% of HR leaders began rehiring within six months
- 17.8% started rebuilding within just three months
- 30.9% found that bringing roles back cost more than the original layoffs saved
This isn't just a pattern overseas. The Commonwealth Bank of Australia walked back its decision to cut 45 customer service roles after deploying a voice bot system. A CBA spokesperson admitted the bank's "initial assessment did not adequately consider all relevant business considerations."
Forrester's 2026 future of work outlook predicts that half of all AI-attributed layoffs will be "quietly reversed." Gartner goes further: by 2027, half of companies that reduced customer service headcount due to AI will rehire for those functions — even if the job titles change.
Why AI Fell Short
The core problem is simple. Over half of surveyed HR leaders said AI "required more human insight than anticipated." More than 20% reported their AI tools underperformed or failed to deliver expected results. Only 21.4% said AI fully replaced the eliminated roles without causing operational problems.
Take call centre workers, who carry an AI risk score of 7.4 out of 10 on our scale — among the highest exposure ratings in Australia. There are 29,400 Australians in these roles. AI chatbots handle password resets and balance enquiries well enough. But the moment a customer is frustrated, confused, or dealing with something outside the standard script, the bot breaks down. That's exactly why CBA reversed its cuts here.
Or consider general clerks — Australia's largest clerical workforce at 286,600 people, with an AI risk score of 7.0. The routine tasks look automatable on paper. But clerical work also involves institutional knowledge, process exceptions, and cross-team coordination that current AI handles poorly. Companies that automated the routine parts quickly discovered how much of the job wasn't routine at all.
Which Roles Are Caught in the Middle
Not every occupation tells the same story. Here's how some of the most affected Australian roles stack up:
| Occupation | AI Risk Score | Employed | What's Happening |
|---|---|---|---|
| Call centre workers | 7.4 / 10 | 29,400 | CBA reversed cuts after voice bot fell short |
| Software programmers | 6.7 / 10 | 195,400 | WiseTech and Atlassian cut deep, but occupation remains in national shortage |
| General clerks | 7.0 / 10 | 286,600 | High routine-task volume makes them a test case for AI — with mixed results |
| Accounting clerks | 7.2 / 10 | 143,500 | Transaction processing is automatable, but process knowledge isn't |
| Bank workers | 5.8 / 10 | 36,400 | CBA investing $90M to reskill rather than replace |
Software programmers are a particularly telling case. AI coding assistants have boosted developer productivity by 40-55% per sprint — WiseTech's CEO declared "the era of manually writing code as the core act of engineering is over." Yet the occupation remains in national shortage with 195,400 people employed and growing at 15.7% over five years. Even after thousands of cuts, demand for software developers still outstrips supply across Australia. The tools are changing the job, not ending it.
What the Government Data Shows
Jobs and Skills Australia released a landmark generative AI capacity study — the first whole-of-labour-market assessment of AI's impact on Australian work. Their central finding backs up what employers are discovering the hard way: generative AI is more likely to augment the way Australians work than to replace jobs through automation.
The study found clerical, administrative, and entry-level roles face the highest exposure. But exposure doesn't equal replacement. It means those roles will shift — more oversight of AI outputs, less manual data entry, more handling of exceptions that AI gets wrong.
The rehiring data lines up with this. The roles are coming back, but they look different. Companies need people who can work alongside AI, catch its mistakes, and manage the work it can't do. That's augmentation, not automation — and it was the government's prediction all along.
The Real Cost of Getting It Wrong
Nearly a third of organisations discovered that the rehire cycle cost more than the layoffs saved. Factor in the recruitment and onboarding of replacement hires, the institutional knowledge that walked out the door, drops in customer service quality during the transition, and reputational damage with remaining staff — and the AI-first approach looks like an expensive experiment.
The most telling finding: 55.1% of organisations that made AI-related cuts never formally discussed reskilling or redeployment before reaching for the redundancy notices. When asked what they'd do differently, 41.2% said they'd completely rethink their approach. Another 50.3% said they'd be more selective about which roles they cut.
One in three HR leaders said they lost critical skills and expertise they couldn't replace. More than a quarter reported that remaining staff lacked the capabilities to fill the gaps left behind.
What This Means for Your Career
If you're in a role with a high AI risk score, the data suggests the transition is slower and messier than headlines make it sound. Companies are learning — at significant cost — that AI is a tool for augmenting work, not a drop-in replacement for experienced workers.
That said, the roles that return won't look the same as the ones that were cut. They'll require more comfort with AI tools, more emphasis on the tasks AI can't handle — judgment calls, empathy, exception handling — and less focus on routine work AI does well.
The smartest move is to understand where your occupation stands. Check your AI risk score at How Safe Am I?, see the full occupation rankings, or compare roles side by side.
The companies that rushed to replace their workers with AI are now scrambling to undo the damage. The question isn't whether AI will change your job — it will. The question is whether your employer will figure that out before or after the layoff notice.