We Laid Him Off. Then We Hired Him Back.

In May, a company redesign at the software firm Syndio eliminated a number of roles. One of the people let go was Jonathan Vidales, a labour economist who had been there five years.

Around the same time, other jobs were redefined and those employees were asked to take on more responsibility and a sales quota. Management assumed they would accept. These were leaner times, jobs were harder to come by, and the pendulum had swung back towards employers. Staying in a beefed-up role with higher earning potential looked like an obvious choice.

Within a few months, several of the people the company had expected to keep resigned. In the push to be more efficient in the age of AI, it lost some of the talent it needed.

The rehire

Vidales came back. In August he applied for one of the newly opened roles and was rehired. On a video call he said, sheepishly, that returning felt like "sneaking back in the house after I got kicked out."

The more uncomfortable detail came next. When his account was reactivated, he found the company's restructuring announcement still sitting in his inbox — and what struck him was how impersonal it read.

To the people who wrote it, the email announced a restructuring. To him, colleagues he had worked alongside had just lost their jobs. Why, he asked, could the company not have acknowledged them and said something kind about the people leaving?

Three in ten employers eliminated positions after implementing AI, only to add those roles back, according to a 2026 study by the staffing firm Robert Half.

The data behind the anecdote

The pattern is not unusual. Three in ten employers eliminated positions after implementing AI, only to add those roles back, according to an April 2026 study by the global staffing firm Robert Half.

AI-attributed cuts peaked in May, when US employers announced 97,000 job cuts and blamed 40% of them on AI. Zillow cut more than 500 people earlier this month, announced the day before earnings, according to the executive coaching firm Challenger, Gray & Christmas.

'AI wishing' and 'AI washing'

The former Lululemon executive Julie Averill described the two failure modes in a New York Times piece this month: leaders who believe they can fix a problem by waving AI at it, and leaders who attribute job cuts to AI-driven efficiency that does not yet exist.

Both produce the same outcome. Companies imagine what AI might make possible and act before understanding how to get there, then discover the capability has not arrived and rehire.

The costs that do not appear in the model

The efficiency case for a restructuring counts salaries removed. It does not count what the Syndio experience illustrates.

The employees who left were not the ones made redundant. They were the ones asked to absorb more work, who read the announcement, drew a conclusion about how the company regarded people, and started looking. Recruiting, onboarding and retraining their replacements consumed the savings.

None of that argues against changing. Syndio's own shift, from a software consultancy towards an AI product business, required it. The argument is narrower and harder to dismiss: the announcement is part of the restructuring, not an administrative afterthought to it, and the people reading it are the ones you are relying on to stay.