Companies Rehire Workers After AI Tools Fail to Replace Them

Companies that eliminated jobs to deploy artificial intelligence are now rehiring workers to fill those same roles. This trend, dubbed the AI boomerang, is forcing firms to reconsider the speed at which they replace human labor with machines.

Microsoft, Barclays Bank, and Kroger announced AI-related workforce reductions this year. These cuts targeted positions deemed redundant or necessary to free capital for technology investment. However, some of these companies have reversed course after AI models underperformed expectations.

Lucy Beaumont, global senior vice president at U.K.-based technology firm SHL, confirmed the shift. Two years ago, the question of whether companies would rehire after AI cuts was hypothetical. Now, the reversals are happening in real time.

Hiring managers in finance, marketing, law, and health reported eliminating positions primarily due to AI or automation. According to Robert Half, these companies later hired again for similar roles. The technology required more human oversight than anticipated and could not replicate institutional knowledge.

Bill Driscoll, senior district president at Robert Half, noted that AI struggles with critical thinking, collaboration, and creativity. These are functions where institutional knowledge is essential. Driscoll said companies are realizing they need people to perform these specific jobs.

Klarna, a Swedish buy now, pay later provider, began recruiting human customer service representatives again last year. CEO Sebastian Siemiatkowski said that the company's reliance on an AI chatbot to cut costs had gone too far. In June, Ford hired or promoted 350 veteran engineers. The automaker's AI tools for quality control did not perform at the expected level.

Betsy Summers, principal analyst at Forrester, said companies often miss the invisible work employees do beyond their formal job descriptions. This includes sensing emotions, holding impromptu meetings, and putting in extra effort when motivated. Brooks Holtom, a professor of management at Georgetown University, added that company procedures for tasks like data processing are often not well documented.

Holtom said organizations are not spending enough time understanding workflows before assigning work to AI. In customer service roles, AI agents frequently need to refer questions to humans who can assess problem severity and respond creatively. Shareholders often prefer the initial cost-saving narrative, leading companies to quietly admit they were not ready for full automation.

Workers offered their old jobs back are asking for more money and clarity on future staffing. Beaumont said returning employees should request additional training and a clear explanation of advancement opportunities. Holtom advised workers to outline a new job description that leverages technology and their own expertise.

Before accepting an offer, Holtom suggested seeking a one-year renewable contract with a buyout clause. This protects workers if the job is terminated in less than a year. Some returning employees may end up in higher-ranking positions that pay more than the jobs they originally left.