Every layoff headline this month has a body count attached and nothing else. TikTok cutting 250 jobs in Nashville. Apple cutting 147 positions in the Bay Area. An AT&T WARN filing out of New Jersey listing 138 names, dated September 15. Read those in isolation and the story writes itself: tech is shrinking. Read the Challenger, Gray & Christmas numbers sitting underneath them, and the story is a lot more specific, and a lot more useful to anyone actually building a headcount plan right now.
Through August, U.S. employers announced 529,914 job cuts this year, 41 percent fewer than the same period in 2025. Total layoff activity is down, not up. Technology is the exception that's eating all the headline space: 155,126 announced cuts through August, 52 percent more than the first eight months of 2025, and 29 percent of every job cut announced in the entire U.S. economy this year — the largest share of any single industry. If you only read that line, the AI-is-gutting-tech narrative writes itself too.
Here's the line most coverage leaves out. The same Challenger report tracked 119,825 planned hires announced through August, 37 percent more than the same period in 2025. Technology companies announced 19,751 of those planned hires, the most of any industry Challenger tracks. The sector generating the largest share of 2026's job cuts is also generating the largest share of its planned hiring.
That's not an industry shrinking. That's an industry churning its own headcount at a speed no other sector is matching. The firm that publishes the numbers reads them the same way.
"Hiring has also increased over last year by 25%, so while AI is shifting the labor market, it is not dismantling it."
— Andy Challenger, Chief Revenue Officer, Challenger, Gray & Christmas, Challenger, Gray & Christmas
The AI-attribution number is more specific than the headlines suggest, too. Challenger says employers cited AI in 116,175 announced cuts through August — about 22 percent of the year's total. But in August alone, AI accounted for only 3,462 of that month's cuts, while restructuring was the leading stated reason at 16,173. The month people are pointing to as evidence of an AI purge was, by the company's own stated reasons, mostly an org-design purge. That distinction matters for the same reason it mattered in Uber's memo two weeks ago: not every layoff you're reading as an AI story is actually one, and treating them as interchangeable will have you solving the wrong problem.
BLS data from the same window backs up the mixed picture instead of resolving it one direction. Nonfarm payrolls added 162,000 jobs in August; unemployment held at 4.1 percent. The information sector lost 23,000 jobs, 8,000 of them in computing infrastructure, data processing, and web hosting. Real, and worth naming.
But BLS's own ten-year projections, updated three weeks ago, have data scientist employment growing 35 percent from 2025 to 2035, adding roughly 95,400 jobs, with information security analysts projected at 21 percent growth and software developers, QA analysts, and testers at 10 percent. You cannot read "AI is eliminating tech jobs" and "BLS projects double-digit growth in data science, security, and software roles over the next decade" as the same forecast. They're not in tension. They're describing different halves of the same workforce.
Zoox is the cleanest single example of how badly the layoff framing can misfire when applied by default. Zoox got pulled into this month's layoff chatter right alongside TikTok, Apple, and AT&T. And there's no reporting to back it up. Zoox is expanding: Reuters reported September 1 that it extended robotaxi testing to Houston and San Diego, bringing its footprint to twelve locations, and it started charging for rides in Las Vegas in August. The company got swept into a layoff narrative purely by proximity to other headlines in the same week, not by anything happening inside Zoox.
"This expansion marks a significant step forward for Zoox and is driven by the insights from our early deployments. This is our year of growth. We are actively implementing learnings to confidently and safely scale our robotaxi service across the country and bring our differentiated experience to even more riders."
— Aicha Evans, CEO, Zoox, The Robot Report
A company in its year of growth does not read as a company in its layoff year.
None of this means the anxiety is misplaced. Pew Research, surveying 5,273 U.S. workers in October 2024, found 52 percent worried about how AI will be used in their workplace, against 36 percent who felt hopeful. That split is earned. Sixty-three percent of those same workers said little or none of their current work was done with AI, meaning most of them were worried about a transition that, for them personally, hadn't started yet. That was almost two years ago; the exposure has climbed since, and the worry has not receded to meet it. That gap, the distance between what people fear and what they have actually touched, is exactly where a badly-timed headcount decision gets made on vibes instead of data.
If you're building a 2027 headcount plan off this year's headlines, the Challenger numbers say something more useful than "AI is cutting tech jobs": tech employers are running the highest-velocity headcount churn of any sector in the economy, cutting and hiring at the same time, often for different roles inside the same company. That's not a signal to freeze hiring. It's a signal to get specific about which roles on your own chart are in the cut column and which are in the hire column, instead of applying one industry-wide number uniformly to every req you own. Apple's own layoffs included software engineering and machine-learning roles, the category the market treats as safest, while the company simultaneously expands its AI push under John Ternus, who took over as CEO on September 1 with that acceleration widely framed as his defining challenge. If Apple can't cleanly separate "roles we're cutting" from "roles we're growing" inside its own AI org, don't assume your chart is simpler than theirs.
The mistake that keeps coming up in our client conversations is treating "tech layoffs" as a single macro signal instead of the noisy sum of a hundred company-specific restructurings, most of which have nothing to do with each other's reasons. Before you freeze a req or greenlight a cut because the sector-wide number looks bad, ask the Challenger-report question about your own org: which roles here are genuinely being commoditized, and which are the ones you'd be hiring for anyway if the headlines weren't this loud. The sector's cutting and hiring in the same breath. Your plan should too, on purpose, instead of by accident.
VC5 Consulting builds headcount plans role by role, not off a single sector-wide layoff number. If your 2027 plan needs the cut column and the hire column sorted out for real, let's talk.