Cloudflare cut 1,100 jobs. Coinbase cut 700. Upwork shed roughly a quarter of its staff. PayPal announced plans to remove 4,760 workers — about a fifth of its workforce — over the next two to three years. All in roughly the past two weeks, all from companies that explicitly cited AI as the reason.
The total for 2026 already sits at 113,863 tech workers across 179 layoff events, more than double last year’s pace at the same point. The polite name for it is “restructuring around AI.” The honest one is that some functions are being deleted faster than they can be replaced.
The harder question is whether AI is genuinely doing the work or whether companies are using it as cover for cuts they wanted to make anyway. Either way, the labor market for white-collar tech workers in 2026 looks nothing like the one of 2022.
The early data points are consistent. Customer support, content moderation, basic engineering tickets, sales development, recruiting coordination, and back-office finance roles are seeing the steepest cuts. These are jobs with structured, repeatable workflows — exactly what current AI agents handle well enough to scale.
The functions still expanding: AI infrastructure engineers, security, applied ML researchers, GTM leadership, and the rare hybrid roles where one person can prompt an AI agent and review its output. The pyramid is shrinking in the middle and widening at the top.
Layoff waves usually come with hiring plans elsewhere. This time, the same companies cutting staff are not net adding bodies. Cloudflare CEO Matthew Prince explicitly told staff the goal is fewer, more productive people. Coinbase framed its cut as a permanent operating model change, not a recession move.
The honest read: AI is good enough that companies are betting they can permanently run leaner. Some of those bets will fail and reverse hiring will follow. Others will stick and become the new normal. Either way, the median tech worker has lost meaningful leverage in compensation negotiations for the first time in over a decade.
Three moves work regardless of where the AI conversation lands. First, build a six-month cash cushion. Tech severance has gotten thinner and runs out faster than people expect. Second, learn the AI tools your team uses well enough to be the person who teaches them — that profile is currently the safest in the job market. Third, document outcomes, not activities. Resumes built on “shipped feature X” beat resumes built on “managed team Y” right now.
Above all, treat AI fluency like spreadsheet fluency in 1995: not optional, not optional, not optional.
113,863 across 179 confirmed events through mid-May 2026, according to public tracking. The total is well over double the same point in 2025.
PayPal’s plan to remove about 4,760 workers is the largest single announcement, although it will be phased over two to three years.
No federal law requires disclosure of AI as a cause of layoffs. Some states are exploring legislation but nothing has passed.
Both. AI is genuinely replacing some structured workflows. It is also being cited as cover for cost cuts that would have happened regardless.
AI infrastructure, security, applied ML, senior product leadership, and hybrid roles that combine human judgment with AI-tool operation.
Negotiate severance, file unemployment immediately, extend health coverage, and build a portfolio of recent shipped work to lead with in interviews.