Tech companies are reconsidering plans to use artificial intelligence to reduce their headcounts after real-world experiments showed that current systems do not always deliver the expected productivity gains and that addressing their failures can impose an additional burden on workers. This does not indicate a retreat from investment in artificial intelligence, but rather a shift in how companies want to benefit from it.
Meta’s Plan to Reduce Teams Is Scaled Back
According to a restructuring plan, Meta had been moving toward cutting the number of employees in some teams by as much as 60%, while assigning a significant portion of tasks to AI agents. But the plan was withdrawn after the expected efficiency levels were not achieved.
According to research conducted by Reuters based on internal company documents, Meta halted additional cuts that had been scheduled after the first round of layoffs. Internal data showed that AI systems had not delivered the expected increase in productivity, while technical problems and the time employees spent resolving them had increased. Later, CEO Mark Zuckerberg announced that there was no plan for another company-wide mass layoff during 2026.
Klarna’s Lesson: Automation Does Not Eliminate the Need for Humans
Swedish fintech company Klarna followed a similar path. The company had previously announced that its AI-powered customer-service system was capable of doing the work of hundreds of employees, and it largely halted hiring. But problems with service quality prompted it to resume hiring workers.
CEO Sebastian Siemiatkowski acknowledged that the company had focused excessively on cutting costs. The model it later adopted keeps AI responsible for routine requests while ensuring that customers can reach a human employee when necessary.
Technical Jobs Are Not Affected at the Same Pace
Hiring data do not indicate that companies are abandoning AI. According to SignalFire’s 2026 State of Tech Hiring report, overall hiring at major technology companies declined by 25% compared with 2019, while software-engineering hiring fell by only 11%. Software engineers’ share of total hiring also rose from 46% to 55%.
By contrast, fields such as design, marketing, and product management recorded sharper declines, while hiring increased in specializations including AI engineering and machine learning. This reflects demand for employees capable of building and operating new systems, not merely using ready-made tools.
What Is Changing in Practice?
These examples suggest that the most immediate effect of AI may not be an immediate wave of layoffs, but slower hiring, particularly for entry-level positions. A study by Stanford’s Digital Economy Lab, based on U.S. payroll data, found no evidence of widespread job losses caused by AI across the economy as a whole. However, it identified weaker employment outcomes for people aged 22 to 25 in occupations most exposed to the technology.
According to the researchers, a significant part of this effect stems from companies hiring fewer workers at the beginning of their careers, rather than laying them off on a large scale. This is an important point for students and people seeking their first jobs, and it also matters to companies that may redesign training and career-progression paths.
certi.news analysis: The actual change is not abandoning AI, but abandoning the assumption that greater use of it means cutting the number of employees by the same degree. The cases of Meta and Klarna show that output quality, system stability, and the ability to handle non-routine cases remain factors limiting full automation. At the same time, hiring data show that demand is shifting toward engineering skills and the ability to work with AI, while open questions about the future of entry-level jobs and the impact on gaining initial experience remain without a definitive answer.