Companies Are Rehiring After Cutting Jobs for AI. The Story Changed Before the Facts Did

Since 2024 the dominant story has been that AI would replace human work soon. Klarna said its assistant did the work of 700 agents. Mark Zuckerberg spoke of an AI that could be a mid-level engineer, “probably in 2025”. Dario Amodei, the head of Anthropic, warned that half of entry-level white-collar jobs could disappear within five years. Some companies acted on that story: Meta cut about 8,000 jobs, 10% of its staff, and in July Zuckerberg said he had miscalculated the timing, with benefits expected within three to six months, according to Reuters.

For more than a year the signals have pointed the other way. Klarna started hiring customer service agents again in May 2025, because the AI was cheaper but the quality lower. The Commonwealth Bank of Australia withdrew a cut of 45 jobs and admitted its assessment was incomplete. A Robert Half survey of about 2,000 US hiring managers finds that 29% or 32%, depending on the outlet, rehired for roles they had cut because of AI, and 44% in finance. These are self-reported answers and a handful of cases, not a proven trend, and I could not find the original report.

I suspect story and facts parted because two interests matched: sellers had promises to place, and employers had costs to cut. Which of the two weighed more, the data cannot yet say.

Meanwhile responsibility has an address. A cut made on a vendor’s promise is a decision of whoever signs it, and the burden of proof sits with whoever adopts, as with any other system.

What happens next cannot be read yet. Meta gave itself three to six months, and Gartner predicts that by 2027 half of the companies that cut customer service staff because of AI will rehire for similar functions under different titles. Those are forecasts, not measurements. The change of direction already says one thing: the earlier story ran faster than the data.