Cut or hire: what the data on AI and 2026 layoffs actually shows

It has become common to attribute every wave of tech layoffs to artificial intelligence. Block cut 40% of its staff, about four thousand people, directly citing AI’s growing ability to perform tasks once done by humans. Meta announced the layoff of roughly eight thousand employees in May, with another seven thousand reassigned internally, as it redirects investment toward AI infrastructure. Nike, Dow, GM and Oracle made similar cuts throughout the year, almost always justified by the same combination of operational efficiency and automation.

The aggregate numbers, however, tell a more nuanced story. Recent projections point to around 500,000 jobs being eliminated in the United States in 2026 because of artificial intelligence, a sharp jump from the prior year, but one that represents just 0.4% of a workforce of roughly 125 million people. The tech sector has already logged more layoffs this year than in all of last year, but a good share of recent cuts hit senior and specialized roles, not just operational functions, which suggests a deeper reorganization rather than a simple cost cut.

The less discussed detail is how small and mid sized companies are behaving, following an almost opposite logic to large corporations. Rather than using artificial intelligence to shrink headcount, many of them treat the technology as a growth lever and still intend to hire, especially in technical roles. That is an important counterpoint to the dominant narrative that automation necessarily means fewer people employed.

For anyone running a company, the right question is not whether artificial intelligence will replace jobs, that is already happening in specific functions, but which group the business itself falls into. Corporations with the capital to invest heavily in AI infrastructure tend to redesign their entire organizational structure around the technology, cutting middle layers and reinforcing engineering teams. Smaller companies, without that same capital available, tend to use the technology as a productivity tool within their existing structure, which changes the decision on hiring, training or cutting staff entirely.

The real risk for 2026 is not ignoring artificial intelligence in workforce management, it is copying a multinational’s decision without considering whether the business’s own context justifies the same move. Well applied automation increases productivity per employee. Headcount cuts without that real productivity gain simply move the cost problem somewhere else in the operation.