Ingka’s global people and culture manager, Ulrika Biesèrt, put the point in corporate language, saying the company was committed to strengthening workers’ employability through “lifelong learning and development and reskilling.” Asked by Reuters whether AI would reduce headcount, she said, “That’s not what we’re seeing right now.”²
That answer deserves some caution. IKEA is not a fairy tale, and Ingka later announced cuts affecting about 800 office-based positions in March 2026 as it sought a simpler and less costly organization.³ But that does not erase the specific lesson from the Billie rollout. In customer service, Ingka chose to ask what its people could become after AI took over the routine work.
Klarna offers the cautionary version of the same story.
In early 2024, the company made one of the splashiest AI announcements in the technology and business worlds. Its OpenAI-powered assistant had handled 2.3 million conversations in a month, about two-thirds of its customer-service chats. Klarna said it was doing work equivalent to that of 700 full-time agents. Average resolution time dropped from 11 minutes to less than two, and the company expected the assistant to add $40 million to profit in 2024.⁴
The numbers were dazzling, but customer service is not a math problem alone. It is also where a company keeps or loses trust.
By 2025, Klarna was adjusting course. It still used AI heavily, but it began recruiting humans again so customers could always get to a person. A Klarna spokesperson offered the line every CEO experimenting with AI should tape to the conference-room wall: “AI gives us speed. Talent gives us empathy.”⁵
Another Klarna line was even better: “AI solves the easy stuff — our experts handle the moments that matter.”⁵
The dividing line is not difficult to see. Nobody needs a human being to recite a return policy, find an order number or answer the same shipping question for the ten-thousandth time. A good bot can do that instantly, in more than 35 languages, at midnight.⁴
But a disputed bill matters. A late refund matters. A denied loan matters. A medical claim matters. A customer who is angry, embarrassed, broke, confused or afraid does not want to be optimized. He wants to be understood.
Klarna’s CEO later acknowledged the tradeoff. When cost became too dominant in the design of its customer-service operation, Sebastian Siemiatkowski said, “what you end up having is lower quality.” He added that investing in high-quality human support was “the way of the future for us.”⁶
That sentence should haunt the AI layoff boom.
Gartner now predicts that by 2027, half of the companies that attributed customer-service staff reductions to AI will rehire people to perform similar functions, often under different job titles. Its analysts emphasized the continuing need for the expertise, empathy and judgment of human agents.⁷
That does not mean AI is a fad. It means a lot of managers are using a powerful tool with a weak imagination.
The principle predates AI by decades. Technology changes, but the management choice does not: treat workers as disposable costs, or give them the information and authority to help build what comes next.
Ricardo Semler understood that long before chatbots. In Maverick, he described taking over Semco, his father’s old-line Brazilian manufacturing company, and turning it into one of the strangest management experiments in modern business. Workers received more financial information. Managers had to earn authority. Employees gained a voice in decisions normally reserved for executives.⁸
When Brazil’s economy was hit by a severe liquidity crisis, Semco did not simply cut labor and hope for the best. Workers and executives studied the numbers together. Employees accepted temporary pay cuts, but management took larger cuts, profit-sharing increased and workers gained more authority over spending.