The tech industry is laying off and hiring at the same time, and the workers landing on the wrong side of that divide are finding the gap almost impossible to cross.
Nearly 79,000 tech workers lost their jobs in the first three months of 2026. That is the highest quarterly figure since early 2024. At the same time, more than 275,000 AI-related job postings are sitting open in the United States alone, and companies from Amazon to Meta are spending billions expanding their AI divisions. These two facts are not in contradiction. They are the same story told from opposite ends of the same workforce.
Almost half of Q1 2026 layoffs, 47.9 percent according to Nikkei Asia, were attributed to AI. But the more revealing statistic sits elsewhere. Research from Bloomberg suggests that roughly half of those AI-attributed layoffs will result in the same roles being quietly rehired, offshore or at lower salaries. Companies are not simply cutting headcount. They are repricing labor, using AI as the justification and the cover.
Amazon cut approximately 16,000 corporate roles in Q1, accounting for more than half of all tech layoffs in the quarter, while simultaneously investing billions into its AI infrastructure and Anthropic. Microsoft eliminated around 9,000 positions, about 4 percent of its global workforce, as it doubles down on Copilot and Azure AI. Block's CEO Jack Dorsey was unusually candid, saying that AI models had become "an order of magnitude more capable" to explain a 40 percent workforce reduction. Pinterest laid off 700 employees, 15 percent of its staff, and framed the move explicitly as making room to hire more AI-skilled talent. The messaging across these companies is notably consistent, and it is not accidental.
What makes this moment different from previous tech downturns is the speed of the skills divergence. Workers with AI competencies are earning 56 percent more than their non-AI peers, according to industry compensation data. An LLM developer commands around $209,000 in base compensation. A QA engineer, a role largely susceptible to automated testing, now sees declining pay. The same industry, the same labor pool, pulling in opposite directions simultaneously.
The Junior Developer Problem
The hardest trend to talk about honestly is what is happening to early-career developers. Entry-level roles have dropped 20 to 35 percent globally over the past year. AI coding tools have absorbed precisely the tasks that used to be how junior engineers built their skills: boilerplate code, basic bug fixes, documentation, simple CRUD applications. Companies that once hired five juniors now hire two mid-level engineers and a Copilot subscription, and get comparable output for less money. The training pipeline for the next generation of senior engineers is being cut off at the source.
Stack Overflow made the tension explicit in a December 2025 piece framing the situation as AI versus Generation Z. Sixty-four percent of Gen Z workers say they are worried about being laid off, compared to 45 percent of millennials. The anxiety is concentrated among people who entered the workforce in the last two years, which is exactly the group that should theoretically be the most comfortable with new technology. The problem is not digital literacy. The problem is that AI absorbed the entry point.
Senior Engineers Are a Different Story
While junior roles contract, demand for senior and staff-level engineers is growing. More than half of current software engineering openings in the U.S. target senior or staff-level candidates. The reason is straightforward. AI tools amplify experienced engineers rather than replacing them. A senior developer using Claude Code or GitHub Copilot can do the work of a small team. The same tools in the hands of someone without systems design experience produce code that breaks in production. The amplification effect only works if there is expertise to amplify.
Several companies learned this the hard way in 2025. They laid off senior engineers to save costs, replaced them with AI-augmented juniors, and spent early 2026 quietly rebuilding senior capacity, often through offshore hiring in Southeast Asia, Vietnam, and the Philippines, where experienced engineers cost 50 to 70 percent less than equivalent U.S. talent. The need for experience did not go away. The willingness to pay U.S. rates for it did.
The Offshoring Acceleration
This is where the AI labor story connects to a broader economic shift that most coverage misses. Remote hiring and skills-first evaluation have made the global talent market genuinely accessible at scale for the first time. A senior full-stack developer in Vietnam or Malaysia earning $3,000 to $5,000 a month delivers comparable output to an American counterpart earning three to four times that. AI productivity tools eliminate the friction that used to disadvantage distributed teams, faster code reviews, clearer documentation, better asynchronous communication through AI-assisted writing.
The net effect is that AI is not just replacing jobs directly. It is enabling a global repricing of technical labor. U.S. tech salaries grew only 0.8 percent year-over-year, down sharply from 3.5 percent in 2023. Meanwhile, 59 percent of tech professionals in the Dice Tech Salary Report said they feel underpaid, a record high, and 47 percent are actively job hunting. The frustration is not imagined. Real wages in tech are effectively flat while the cost of not having AI skills rises every quarter.
What the Numbers Actually Say About the Future
The World Economic Forum projects 170 million new AI-related jobs globally by 2030, against 92 million displaced. That is a net gain of 78 million positions. The U.S. Bureau of Labor Statistics projects 17 percent employment growth for software engineers through 2033, representing roughly 328,000 new jobs. BCG estimates that around 12 percent of current jobs will face outright substitution by AI. The remaining 88 percent will be reshaped rather than eliminated.
None of that is particularly comforting to a laid-off senior developer at 44 who specialized in frameworks that AI now generates automatically. The macro numbers are positive. The micro experience of being on the wrong side of this transition is a different matter entirely. The labor market is not broken. It is recalibrating. But recalibration has always required real people to absorb the cost of the adjustment, and this round is moving faster than any retraining infrastructure can match.
The companies that understand this are already positioning for the next phase. Hiring AI-fluent engineers at every level, investing in internal upskilling rather than pure headcount reduction, and thinking about global talent pools not as a cost-cutting mechanism but as a structural advantage. The ones treating AI as a one-time opportunity to slash the payroll budget will find themselves short of the senior talent that actually makes AI work when the next product cycle demands it.
Also read: Caltech's PrismML shrinks AI models to fit your phone without losing their mind • Z.ai's GLM-5. and GLM-5.1 discount looks like a defense against DeepSeek pressure • India's new AI takedown rules could restrict free speech faster than deepfakes
