Some employers have discovered the hard way that firing people because ”AI can do it” is not the same thing as running a business. Ford, Commonwealth Bank of Australia, and IBM are all examples of companies that leaned into automation, hit the limits of the machines, and then moved back toward human hiring when the work got messy, high-volume, or simply too nuanced for software.
The pattern is awkward but unsurprising. Productivity tools are great at trimming routine tasks; they are much less impressive when customers start calling back, quality problems pile up, or a tiny slice of edge cases turns into a giant headache. The real winners in this round are employees with hard-to-automate judgment, while the losers are the managers who mistook partial automation for full replacement.
Ford, CBA and IBM reverse course
Ford is hiring hundreds of experienced engineers again to deal with quality issues that automated systems failed to solve. In Australia, Commonwealth Bank of Australia cut more than 40 customer service jobs after replacing them with a voice AI bot, then had to walk that decision back when call volumes rose instead of falling. IBM went in a different direction but reached a similar conclusion: its AI handled about 94% of routine HR requests, yet the remaining 6% included ethical questions the software could not answer well, so the company said it plans to triple entry-level hiring in the US across all business units in 2026.
That last move matters because it undercuts a familiar executive fantasy: cut headcount first, figure out the workflow later. IBM’s own experience suggests that the awkward, human part of work is often the part you still need most once automation starts flagging. And that is before you get to the cost of rework, customer frustration, and the time spent explaining why the ”efficient” system created more noise than value.
The numbers behind the AI hiring reversal
The cautionary data is piling up. Orgvue found that 39% of business leaders reduced staff because of AI, and more than half, 55%, later said that was a mistake. Robert Half reported that 32% of hiring managers in the US cut vacancies because of AI and then hired people back into the same or similar roles. That is not a minor course correction; it is a very expensive admission that the spreadsheet moved faster than the strategy.
- 39% of business leaders cut staff because of AI, according to Orgvue.
- 55% of those leaders later said the cuts were a mistake.
- 32% of US hiring managers told Robert Half they reduced openings due to AI and then rehired for the same or similar jobs.
Why automation alone keeps failing
Intuition Labs puts the problem plainly: companies budgeted for ”technology to replace people” without spending enough on training or upskilling, which left teams unprepared to work with AI instead of underneath it. In practice, that means organizations often remove the very staff members who could supervise the systems, catch mistakes, and handle the exceptions that determine whether customers stay calm or leave in a huff.
That is why the more sensible model is emerging now: humans plus AI, not humans versus AI. The next wave of hiring is likely to favor workers who can audit outputs, handle edge cases, and translate machine output into something customers can actually live with. If executives want to save money, they may need to stop trying to prove that a bot can do everything and start asking what the bot still cannot do well enough.