Questions we get asked

It depends on which count you look at, and we keep them apart on purpose. When we last checked Layoffs.fyi (September 10, 2026), its industry-wide count stood at 128,536 people across 299 tech companies. That's already past the 122,606 it counted for all of 2025. Our own tracker is smaller by design, because it only includes major events we can source one by one (the live numbers are on the tracker page). The biggest single event of 2026 in our data is Oracle's 21,000-position restructuring. The makeup of the cuts has changed too. Most of the people laid off this year worked at companies that explicitly cited AI, and most of the companies cutting are profitable, so we call this restructuring rather than distress. And layoffs are only half of it. Hiring intent and software postings both rose through 2026, which is why we publish the bright spots next to the cuts.

Yes, and more than the headlines would have you think. In Robert Half's survey, 78% of technology leaders plan to add permanent headcount in the second half of 2026, up from 61% at the start of the year. Indeed's software-development postings index is 22% above its May 2025 low, and 6.3% of all US job ads now mention AI, nearly double the 2022 peak. The demand is lopsided, though. Experienced specialists in AI/ML, data, security and platform work are doing best. Financial services and manufacturing are hiring more than big tech, and startups and mid-market software companies more than the giants. So we'd describe the market as sorting people more than shrinking. Our Opportunities page tracks where it's growing, with sources.

Both, depending on where you're standing, so be wary of anyone who sums it up in one mood. On the bad side, 2026 layoffs have already passed 2025's total, entry-level hiring is still depressed, generalist pay has come down from the 2022 peak, and AI is now the reason companies cite most often for cuts. On the good side, software postings have been climbing for more than a year, hiring intent is the highest it's been since 2022, the wage premium for AI skills has widened to 56%, and the AI labs, banks, manufacturers and health systems are all hiring more people than they let go. If you're an experienced specialist who's comfortable with AI, you're in a seller's market. If you're junior or a generalist, you'll probably need to reposition, and that's what our skills pages and the resilience calculator are for.

Because most of today's cuts are restructuring, and being profitable doesn't stop that. Part of it is still the hangover from 2020 and 2021, when big tech grew headcount by 40 to 60% in two years. Part is investor pressure. Once the 2022 rate hikes ended the free-money era, Wall Street started rewarding 'efficiency', and a layoff announcement now tends to come with a stock bump. The newest piece is AI reallocation, where a company cuts jobs in one place to pay for AI hiring and data center spending somewhere else. When a profitable company trims 5% in the same quarter it posts record earnings, it's rebalancing its bets, not fighting for its life.

For visible, remote-friendly postings, yes, and it happens all the time. LinkedIn's own data and recruiter surveys through mid-2026 show popular postings pulling in 300 to 1,000+ applications within days. Part of the reason is AI tools that let candidates mass-apply (some people send thousands of applications). That floods every opening and pushes employers toward AI screening, and the resulting arms race has made applying through the front door close to useless. We don't think that's a reason to despair. What works now is referrals, direct outreach and being easy to find through a portfolio, open-source work or things you've written, and all of those do far better than cold applications.

In 2020 and 2021, tech companies assumed pandemic demand for e-commerce, streaming and remote tools would stick, and they hired to match. Meta nearly doubled. Amazon added about 800k roles. Zoom, Peloton and Shopify scaled up for a world that partly went back to how it was. When demand cooled and interest rates rose in 2022, that extra headcount was the first thing to go. It still matters because it taught companies and investors that big layoffs are survivable, even rewarded, which made the AI-driven cuts that came next a lot easier to justify.

From 2022 to 2025, the hardest-hit functions were recruiting and HR (hiring stopped, so the people who do the hiring went first), customer support (the first target for AI agents), manual QA, routine content production, data entry and back-office operations, middle management (the 'flattening'), and entry-level roles across the board. The best protected have been security, AI/ML engineering, data engineering, senior engineers who own systems, and anyone who pairs technical skill with expertise in a regulated field.

Both, and the balance is shifting. In 2022 and 2023, 'AI' was mostly a cover story for correcting over-hiring. By 2024 and 2025 the direct cases were hard to argue with: Duolingo's translators, Stack Overflow's traffic, Chegg's business model, Salesforce's 4,000 support roles, IBM's unfilled back-office positions. The bigger effect gets less attention, though. It's the jobs that never get created. A team that would have hired 10 people hires 6, and the missing 4 never show up in any layoff tracker. That's why we think hiring-rate data tells you more than layoff counts do.

Think in tasks. Job titles won't tell you much. Work holds up when it needs someone physically present and good with their hands, or when a person has to carry legal accountability for it (medicine, engineering sign-off, audits). It also holds up when it calls for high-stakes judgment in murky situations, depends on deep relationships and trust, or sits at the frontier where there isn't much training data. Work is exposed when it's routine, digital, clearly specified and easy to check, however skilled it used to be. A rough test we use: if you could hand the task to a very fast, well-read intern who carries no accountability, AI will eventually do it.

Probably, but the timing is the problem. In past technology shifts, specific jobs disappeared fast and new categories showed up slowly, usually with a gap of 5 to 10 years, and the people who lost the old jobs were rarely the ones hired into the new ones. You can already see new titles appearing, like AI operations, agent orchestration, evals and AI governance. Our honest answer is that the totals may balance out, but your own outcome depends on whether you move toward the new work or wait for the old work to come back. We don't expect it to.

Yes, though coding has become the entry fee rather than the thing you're selling. AI writes code now, so the value has moved to everything around it: deciding what to build, breaking problems down, reviewing and debugging what the model produces, system design, security, and actually shipping a finished product. If you learn to code with AI, understanding what you're doing but moving faster, you're building an advantage. If you learn by memorizing syntax and doing slowly what AI already does, you're preparing for 2019.

As a foundation, yes. As a job guarantee, no, and that's the part that changed. New-grad hiring has roughly halved since 2019, and degree holders now compete with bootcamp grads, self-taught AI-native builders and each other. The degree still helps with visa pathways and big-company filters, and the fundamentals matter more now that AI writes the easy code. Just don't treat it as your whole case. Employers now consistently rank proof that you can ship, meaning real projects with real users, above credentials.

Noticeably longer than in 2021, when offers came almost instantly. Surveys through mid-2026 put a typical search at 3 to 6 months, often longer for senior and niche roles, and new grads face the longest odds. Budget for 6 months. Run the search like a sales pipeline, with lots of threads going at once instead of one application after another, and expect the offer that lands to come through a referral or someone you know, far more often than through a job portal. If you still have a job and you're worried, start looking before you need to. Employed candidates still convert much better.

Start with the paperwork: severance terms, equity deadlines (you often get just 90 days to exercise options) and health coverage. Don't sign anything on day one. File for unemployment right away, because that's exactly what it's for. Then give yourself a week to process it, since panic-applying produces bad applications. While memories are fresh, line up 3 to 5 references and ask for LinkedIn recommendations. After that, run a structured search. Pick 2 or 3 target role types, go to your network before you apply cold, and give your days a routine. Try to see the layoff as a market event rather than a verdict on you. From 2022 to 2025, half a million skilled people were laid off, and hiring managers know it.

Sometimes, and asking (politely) costs you very little. More is negotiable than most people assume. You can push on your departure date (to reach a vesting cliff), on how your equity is treated (accelerated vesting or a longer exercise window), on healthcare continuation and outplacement support, and on whether you're listed as eligible for rehire and a reference. The cash amount is the hardest thing to move in a mass layoff, because companies worry about setting a precedent. Your own situation can still give you room to push, for example long tenure, recent high ratings or a protected-class concern. If the numbers are large, an hour with an employment lawyer usually pays for itself.

The strongest demand is for building with AI: LLM integration, RAG, agents and evals. Using ChatGPT doesn't count. After that come the skills with steady premiums, which are security, cloud and platform engineering, and data engineering. Then there are the multipliers that make everything else worth more, like depth in a regulated industry, communication and stakeholder skills, and a public track record. Everything on that list does one of four things. It builds AI systems, secures them, feeds them data, or supplies the judgment they lack.

Think of it as four levels. At the first, AI is your assistant for drafting, summarizing, and explaining code or ideas you don't know yet. At the second, it's built into your workflow, with standing prompts and templates for recurring tasks and AI in your IDE, inbox and docs. At the third, you're automating: scripts and agents finish whole tasks (report generation, data cleaning, a first-pass code review) and you edit the result. At the fourth, you're shipping AI features to customers. Fewer people get to each level up, and each one pays off more. Keep a record of what you automate, because 'I built an AI workflow that saved my team X hours a week' is a resume line that works in any function.

Right now the data says experience protects you. Layoffs have hit junior and high-volume roles hardest, and mid-career people bring what AI lacks, such as judgment, domain context and a feel for how organizations really work. The real risk for experienced people is a different one. You can get too attached to a way of working that's ending. The change you need to make is usually smaller than you fear, since your expertise is still what you're selling and AI mostly changes how you deliver it. A 45-year-old accountant who runs AI-assisted close processes is more valuable than the old version of that accountant, and more valuable than a junior with prompts.

We build the layoff tracker by hand from primary sources: company announcements, SEC filings and credible reporting (Reuters, Bloomberg, CNBC, the Financial Times and outlets like them). Every event links to its source, and the counts are what was reported when the cut was announced. We're conservative with the AI flag and only apply it when the company itself or credible reporting tied the cut to AI. For wider context we use public research from Challenger Gray, LinkedIn's Economic Graph and layoffs.fyi. Keep in mind that we track announced events, and the final numbers sometimes turn out different.

It's a weighted scoring model, and we publish every part of it. It looks at your role's task-level exposure to AI (based on published research on task automation), current hiring demand for the role, how stable your industry is, your experience level, how much of your work is routine, how fluent you are with AI, and protective skills like domain depth, leadership and security. All of the weights are listed in the methodology section of the calculator page. It can't predict your future. Use the score as a structured way to think about where you stand, and pay more attention to the recommendations than to the number.

Yes. The tracker, the calculator, the skills guides, the predictions and every article are free. We run it this way because working out what to do in a job market like this one is hard enough without a paywall between you and the information. If you find it useful, the best thing you can do is pass it to someone whose career it might help.