The Tech Job Market in 2026: What the Numbers Actually Say
By mid-June 2026, more than 153,000 tech workers had been laid off this year according to Computerworld’s running 2026 tracker, with Oracle cutting 30,000 alone. That number sits alongside a genuinely confusing counterpoint: top tech companies are also hiring 20% more software engineers compared to a year ago, according to Gergely Orosz’s 2026 job market analysis at The Pragmatic Engineer. Both things are true at the same time. That’s the actual state of the tech job market right now, and it’s why blanket advice (“hiring is back!” or “SWE jobs are dying!”) is so consistently useless.
I’ve been talking to a lot of engineers navigating this. The honest answer is that the market in mid-2026 depends almost entirely on which category you’re in. Not which company. Which category.
What the structural shift actually looks like
The BLS projects software developer employment to grow 15% from 2024 to 2034, which sounds reassuring. It adds roughly 288,000 new positions over that decade. The median annual wage for computer and IT occupations hit $105,990 in May 2024. On those metrics, the career path is still sound.
The problem is that the growth isn’t spread evenly. It’s concentrated. Entry-level software engineering postings are down roughly 28% from 2022 peaks. At the same time, AI/ML engineering job postings were up 85% year-over-year in recent tracking, and those roles have pushed past 50% of new tech job listings overall – up from about 10% in 2023. The market didn’t shrink. It reorganized.
This is, of course, not great news if you’re a generalist mid-career engineer who hasn’t worked directly with ML systems. Many engineers I know are in exactly that position. Some are finding the transition fine; others are finding it a lot harder than they expected. Both experiences are real.
Where hiring is actually happening
Fintech and security are expanding fastest right now. Ramp grew engineering headcount by 94%, Wiz by 84%, Datadog by 68% – all per the Pragmatic Engineer data. These aren’t household names in the same way as the FAANG companies, but they’re paying competitively and actually hiring.
Big tech is messier. Apple is up roughly 10% in headcount over a two-year window. Google is up 5%. Microsoft and Amazon are slightly negative over the same period, though both are actively hiring in specific AI infrastructure roles while contracting elsewhere.
The hiring windows matter here. The Pragmatic Engineer data points out that most net hiring in software engineering happens between March and June. If you’re job searching in August, you’ll feel the market is tighter than it actually is – it’s partly seasonal, not purely structural.
The AI skills question is real, but messier than it sounds
The 2025 Stack Overflow Developer Survey found that 84% of developers use or plan to use AI tools, up from 76% in 2024. 51% use them daily. But here’s the part that often gets cut from the summary: only 33% trust AI output to be accurate, and 46% actively distrust it. Experienced developers are the most skeptical – 20% report “high distrust” of AI accuracy.
What this means practically: employers in 2026 want engineers who can work with AI tools and who understand their limits well enough to catch the mistakes. That’s different from just wanting engineers who can prompt ChatGPT.
Whether companies will eventually decide AI agents can fully replace mid-level software engineers is genuinely uncertain. I don’t think anyone has a confident answer to that. The people who claim they do are usually selling something.
Skills that are moving the needle in interviews
- AI/ML engineering (model fine-tuning, RAG pipelines, inference optimization)
- Platform and infrastructure work – especially anything touching LLM deployment at scale
- Security engineering, which has seen consistent demand growth separate from the AI wave
- Full-stack development with real AI integration experience (not just “used Copilot”)
- Data engineering, particularly around training pipelines and feature stores
The skills that look weaker on resumes in 2026 aren’t disappearing – CRUD apps still get built, SQL still matters – but they’re table stakes, not differentiators.
What we notice in mock interview practice
Candidates who practice on LastRoundAI’s mock interview tool tend to get caught by the same behavioral question in 2026: “Describe how you’ve used AI tools in a production context.” The engineers who can give a specific, concrete answer – real system, real trade-off they navigated, real failure they caught – perform noticeably better than those who answer generically. It’s not a gotcha question; interviewers just want evidence that the AI skills on the resume are real.
The interview process has genuinely changed
Two patterns that weren’t common in 2022 are now standard. First, take-home assignments have gotten longer and more ambiguous – sometimes 4 to 6 hours of work, with evaluation criteria that aren’t clearly spelled out. Second, behavioral rounds now carry more weight in final decisions, partly because with AI assistance, coding round performance has become harder to differentiate.
A lot of candidates I talk to underinvest in the behavioral prep because they assume technical performance is what moves the needle. That was probably true in 2020. It’s less clearly true now.
Practicing behavioral interview questions and AI screening interviews is worth more time in 2026 than it used to be. And if you’re aiming at AI/ML roles specifically, the machine learning interview question patterns have shifted toward practical deployment scenarios rather than pure ML theory.
What the market probably looks like through end of 2026
The BLS 15% growth projection covers a decade, so it doesn’t tell you much about the next six months. Short-range, the layoff pace so far in 2026 (about 1,000 people per day by some tracker estimates) suggests continued restructuring at large companies. But the fintech, security, and AI infrastructure segments are absorbing a lot of displaced engineers.
Entry-level hiring remains genuinely tight. That’s the honest part. Senior engineers with AI system experience are, by most accounts, getting multiple offers relatively quickly. The middle, the engineers with 4 to 7 years of experience who haven’t made a deliberate push into AI tooling, are having the hardest time. That’s a guess based on patterns I’ve observed, not a data point I can fully back up.
What I’m fairly confident about: the companies doing targeted, skills-based hiring rather than headcount hiring are the ones worth your time. Interview processes at those companies are longer and more demanding. Preparation matters more than it did when hiring was hot and undiscriminating.
