Tech Job Market 2026: More Layoffs Than 2025, and More Postings

On September 10, Layoffs.fyi's count of tech workers cut this year reached 128,536, spread across 299 companies. That's already more than the 122,606 it logged for the whole of 2025, and there are almost four months left on the calendar.

Now pull up a different chart. Indeed's index of software-development job postings sat at 61.1 in May 2025. In August 2026 it was 74.4, and it hasn't had a down quarter in between.

Both numbers are right. They count different things, and we think the gap between them tells you more about the tech job market in 2026 than either one does on its own. The layoffs are piled up at a few dozen large companies rearranging themselves around AI spending. The postings come from thousands of employers, most of which would never call themselves tech companies. If the layoff chart is the only one you look at, you'll end up aiming your search at the part of the market that's getting smaller.

The layoff count passed 2025 in September

This year's cuts are bigger than last year's, and they've started to feel routine. Oracle's 21,000-position restructuring in March is still the largest single event. Meta cut 8,000 in May and said plainly it was making room for AI spending. Even Apple, which sat out the entire 2022 to 2025 cycle, cut just over 200 roles from its Siri and Vision Pro teams in August.

Then came Uber. In September it removed 3,300 roles, 10% of its staff, with a memo that read more like an org-design spec than a layoff notice: 20% fewer managers, half as many one- and two-person teams, and nobody more than seven layers below the CEO.

The reasons companies give have shifted as well. Challenger, Gray & Christmas counted 116,175 announced US job cuts through August that named AI as a factor, the first year AI has topped its attribution table. Our own tracker uses a stricter test. We only mark an event as AI-attributed when the company or credible reporting made the link, and even then, most of the people cut this year worked at companies that cited AI.

What you won't find in the ledger is distress. After the cut, Uber's headcount is roughly back to its 2021 level, on far higher revenue. Chime cut 10% five days before it reported a profitable quarter. These companies are moving money from one budget to another, which is why the cuts and the hiring can happen in the same year.

Tech leaders say they're adding staff

Hiring data moves slowly and doesn't come with a villain, so nobody screenshots it. Robert Half's mid-year survey found 78% of technology leaders planning to add permanent headcount in the second half of 2026, up from 61% at the start of the year, and 66% also planning more contract hires. Two thirds said skilled people were harder to find than a year earlier. You don't hear that sentence in a market that's collapsing.

Indeed's August snapshot fills in the demand side. Software-development postings are 22% above their May 2025 low. Job ads that mention AI reached 6.3% of all US postings, close to double the 2022 peak of 3.3%, and much of that growth sits outside the labs, in ordinary product, marketing and operations jobs that now ask for AI skills. The wider economy is helping: 162,000 jobs added in August, with unemployment steady at 4.1%.

The loudest hiring is at the AI labs. OpenAI told staff it aims to nearly double headcount to around 8,000 by year-end, mostly in engineering, research, product and enterprise sales. Anthropic's open roles roughly doubled year over year. Big tech as a group is a net cutter this year. The companies building the technology are net hirers.

The cuts and the hires are different jobs

Our read is that the market is sorting people rather than shrinking. The companies doing the cutting are taking out coordination layers and the support work AI now absorbs, and at the same time they're hiring for the jobs that build AI and the ones that keep it secure and governed. Intuit did this openly in 2024, when it cut 1,800 people and hired 1,800 different ones. This year looks like that move repeated across a whole industry.

Where the hiring happens is different too. The postings rebound is concentrated in startups, mid-market software, financial services and manufacturing. It isn't coming from the ten most famous employers, so a search pointed only at big tech is pointed at the part of the market that's trimming.

We should be honest about where the evidence is soft. Postings measure intent, and a survey of tech leaders measures plans. Neither one is a hire, so treat the hiring side of the ledger as a strong hint rather than proof.

Experienced specialists are getting the offers

If you're an experienced specialist, the market is better than the headlines. AI/ML engineers, data engineers, security engineers and platform engineers top every in-demand list, and Robert Half's starting ranges for them run from $118,500 to $193,250. PwC measured a 56% wage premium for roles that require AI skills, up from 25% a year earlier, comparing the same job titles.

If you're junior or a generalist, it's worse. Employers say they want someone who can own an AI workflow redesign on day one, and 'AI adoption and automation workflows' is the single capability they report as hardest to hire. Entry-level postings are still depressed. Generalist software pay has reset from the 2022 peak. And the whole-team cuts of 2026 fall hardest on people who don't have an internal sponsor.

So the first decision is working out which of those two readers you are.

If you're on the shrinking side, the job for this quarter is to move yourself across, and none of it needs a new degree. Point your search at the employers adding people, such as banks, manufacturers, health systems, mid-market software and the AI labs, instead of only at the giants doing the cutting. At your current job, find one workflow you can redesign around AI and write down the hours it saved. It's the capability employers say they can't find, and it travels well on a resume.

Generalists should pick the nearest specialist path with real demand behind it: data engineering, platform, security or AI product. Our role comparison tool shows the exposure and demand gap for any pair. Juniors will do better with residency-style programs and non-tech employers hiring their first AI-native cohort, because that's where entry-level hiring is still happening.

We keep both sides of the ledger on the site, the tracker for cuts and the Opportunities page for hiring, each with sources and dates and refreshed monthly. The thing to watch this fall is whether Indeed's index keeps its streak while the layoff count keeps climbing. If postings roll over, the two-chart story stops being reassuring, and we'll tell you.