Tech Layoffs In 2026 Odds: The AI Model Verdict Vs. The Kalshi Market
Few economic stories have carried more standing search demand this year than a blunt one: are the tech layoffs getting worse? Kalshi runs a contract on exactly that, asking whether 2026 will see more layoffs in the tech sector than 2025 did. Traders have a firm answer. The market trades near 90 cents, treating "yes" as close to a done deal. We put the same question to six AI models without letting any of them see that price. They came back even more sure than the crowd, and the reason is a number that is already on the board.
The Quick Answer
Kalshi's "More tech layoffs in 2026 than in 2025?" contract trades near 90¢ (about 90% implied). Our price-blind, six-model panel lands higher, at a 96% blend. The reason is arithmetic that is already most of the way settled: through May, the settlement series had logged 281,000 layoffs against a full-year bar of 447,000, so the final seven months need only about 23,700 a month to clear it, versus a 56,200-a-month pace the year has actually run. For a "no," layoffs would have to more than halve and stay there.
- Market Price: near 90¢ YES (~90% implied), as of August 3, 2026
- AI Blend: 96% YES (median ~96.5%) across six models
- The Key Fact: 281,000 layoffs already booked through May 2026, about 63% of the whole 447,000 bar
- The Bar: the full-year 2026 total must top 447,000, which was 2025's exact total
- The Catch Worth Knowing: this settles on the whole BLS Information sector, not just Big-Tech headlines, and on the FRED figure as later revised
Why the panel sits above the market, where the sliver of residual doubt actually lives, the month-by-month settlement data, and a plain-English guide to what this contract really pays out on are all below.
The Market At A Glance
| Market | More tech layoffs in 2026 than in 2025? |
| Venue | Kalshi, a CFTC-regulated event-contract exchange (18+; availability varies by state, as of August 2026) |
| Ticker | KXLAYOFFSYINFO-26 (single YES/NO contract) |
| Settles | YES if total 2026 layoffs in the BLS Information sector exceed 447,000, per FRED series JTU5100LDL. Resolution follows the finalized December 2026 JOLTS data; the contract's expiration is set for March 1, 2027. |
| Market YES Price | near 90¢ (~90% implied) |
| Panel YES Blend | 96% (six models) |
| Prices As Of | August 3, 2026 |
These are model estimates, not predictions of fact and not financial advice. Kalshi event contracts trade on a CFTC-regulated exchange; you must be 18 or older and in an eligible state to participate. The price above is read from Kalshi's live quote and refreshes as the data develops; the "as of" date carries the timestamp.
What This Market Actually Counts
Two things about the settlement rule matter before you read any number on this page, and both are easy to miss from the title.
First, the word "tech" is a label, not the measurement. The contract's own settlement rule reads:
"If there are more than 447,000 layoffs in the information sector in 2026, then the market resolves to Yes."
The Information sector is a formal U.S. Bureau of Labor Statistics grouping (NAICS code 51). It includes software publishing and traditional publishing, telecommunications, broadcasting, motion pictures and sound recording, and data processing, web hosting, and related services. So this is a bet on layoffs across the whole information economy, tracked through the JOLTS survey and published as FRED series JTU5100LDL, not a headcount of the marquee Silicon Valley names that make headlines. That breadth actually stabilizes the "yes" case, because it does not depend on any single company's news.
Second, the bar is 447,000, and there is a story behind that exact figure. The market was originally listed with a 494,000 underlying, which Kalshi later flagged as an error: the true 2025 Information-sector layoff total was 447,000, and the rulebook was corrected to that number. Strip the housekeeping away and the live question is clean: will 2026 book more Information-sector layoffs than 2025's 447,000? The panel scored it on exactly that bar.
The Number That Is Already On The Board
Here is the fetched ground truth the whole question turns on. These are the FRED JTU5100LDL monthly figures (layoffs and discharges in the Information sector, in thousands), the exact series that settles the contract, through the latest reported month.
| Year | Full-year Information-sector layoffs | Note |
|---|---|---|
| 2024 | 395,000 | the prior baseline |
| 2025 | 447,000 | up 13% on 2024; this is the bar |
| 2026 (Jan–May) | 281,000 | five months, and already ~63% of the bar |
The 2026 months, one by one, tell the story: January 71,000, February 42,000, March 60,000, April 56,000, May 52,000. That is 281,000 booked before summer even starts, at an average pace of 56,200 per month. For comparison, the same five months of 2025 produced 177,000, so 2026 is running 59% ahead year over year in the very series that decides this market.
The direction is as important as the level. Annual layoffs in this sector rose from 395,000 in 2024 to 447,000 in 2025, and 2026 is tracking well above both. This is not a market asking whether a stable trend might tick up; it is asking whether a rising trend will keep rising by any margin at all.
The Math To YES
This is where the market's high price stops being a mood and becomes a calculation. The contract needs the full-year 2026 total to clear 447,000. Do the subtraction:
- Through May, 2026 has already logged 281,000.
- That leaves 166,000 to be added across the final seven months, June through December.
- 166,000 over seven months is about 23,700 per month.
Now measure that required pace against reality. The 2026 run-rate is 56,200 a month, more than double what is needed. Every single month of 2026 has printed at least 42,000. Every month of 2025 printed at least 31,000. The lowest monthly figure anywhere in the last three years is 20,000, back in February 2024. A straight-line extrapolation of the current pace lands the full-year 2026 total near 674,000, roughly 50% above the bar.
Put plainly: for this market to resolve NO, Information-sector layoffs would have to fall to less than half of the pace they have run all year and stay there for seven straight months, undercutting even the softest single month in the entire dataset. That is why the panel is not near the market's 90 cents. It is higher.
The Case For YES: Why The Panel Sits Above The Market
The bull case here is not a forecast so much as an accounting fact with five months of runway already removed from it. More than 60% of the target has already been recorded and cannot un-happen; revisions can nudge it, but the bulk is banked. The remaining bar is so low relative to the trend that even a meaningful slowdown clears it. As one seat noted, even if the back half of 2026 ran at 2025's single weakest month (November's 31,000) every month, that alone would add 217,000, blowing past the 166,000 needed with room to spare.
The breadth of the settlement series reinforces it. Because the contract covers the entire Information sector rather than a shortlist of tech giants, a "yes" does not need another dramatic wave of Big-Tech cuts. Structural churn across telecom, media, publishing, and data services provides a floor under the monthly numbers regardless of whether the next viral layoff headline ever arrives. When the required pace is 23,700 a month and the sector's structural baseline sits well above that, the arithmetic does most of the work.
The Case For NO: What Would Have To Break
An honest page names the other side, even when it is thin. The market itself keeps roughly 10 cents on the NO side, so the crowd is not calling this a certainty, and there are two ways it gets there. The panel weighed both.
The first is a sudden, sustained collapse in hiring-and-firing activity. If the elevated early-2026 prints, especially the outsized 71,000 in January, reflected one-time restructuring waves that simply exhaust themselves, the back half of the year could decelerate sharply. Layoff series do carry some seasonality, and a sharp labor-market turnaround, a rapid end to tech restructuring paired with a spending revival, could in theory pull the monthly average down toward the low-20,000s. It has never happened in this window, but "never in three years" is not "impossible."
The second is a data revision. Settlement keys on the FRED figure as revised, and JOLTS numbers do get adjusted after first publication. A large enough downward revision to the already-reported 2026 months could, in principle, lower the running total. The panel judged a revision big enough to matter, on the order of tens of thousands across five months, to be historically rare. Both of these are why the blend sits at 96% rather than at a false 99%, and why the single most-watched data point from here is the next monthly JOLTS print: another 40,000-plus month would make the question academic, while a shock reading below 25,000 would be the first real evidence for the collapse scenario.
Model Verdicts
Market: about 90% · AI blend: 96% (median ~96.5%). The panel confirms the direction and pushes past the crowd's 90¢.
| Model | YES probability | Why |
|---|---|---|
| Claude Fable | 97% | 281K already booked; the final seven months need ~23,700/month against a 56,200 pace, below every single 2025 and 2026 monthly print. |
| Claude Opus | 97% | A NO requires layoffs to halve and hold for seven straight months, below every 2025–2026 print, from a trend that is rising, not stable. |
| Claude Sonnet | 96% | The remaining bar is less than half the run-rate; only a large data revision or an unprecedented sector-wide freeze flips it. |
| GLM | 95% | Even 2025's weakest month repeated seven times clears the gap; the trend is actively rising, though back-half seasonality is the one real caveat. |
| Kimi | 95% | NO demands a ~38–40% year-over-year collapse starting immediately after a stretch running 59% ahead; layoff series rarely halve overnight. |
| DeepSeek | 98% | Already 62.9% of the bar with seven months still to run; even a downward revision leaves a trivial hurdle, and only a broad-based hiring surge threatens YES. |
| Blended Verdict | 96% | equal-weight mean of six models (method always disclosed); median ~96.5% |
Model estimates generated August 3, 2026, price-blind, then passed through a second round where each model read the others' anonymized reasoning and could revise. These are model estimates, not predictions of fact and not financial or trading advice. Models are frequently wrong; the market price reflects real traders' money.
The revision round did little work here, which is itself the finding. Only GLM moved, nudging from 93% to 95% after a peer stressed that the annual trend is actively rising rather than merely holding steady. Every other seat held firm, each noting that the peers had run the identical arithmetic and surfaced no new evidence. It is worth being honest about one limit of that agreement: these are large language models drawing on overlapping training data, so six of them clustering in a three-point band is partly shared priors, not six fully independent reads. Treat the consensus as a soft confirmation of the accounting, not as proof.
Where The Residual Doubt Lives
There is no real bull-bear split on this board; the widest gap on the panel is three points, from GLM's 95% to DeepSeek's 98%. That in itself is the story, because the panel disagrees plenty on murkier markets. Here the numbers are already too far along to argue about. GLM is the closest thing to a cautious voice, and its caution is narrow and specific.
"Even the single lowest back-half month of 2025 (November at 31,000) sustained across all seven remaining months would yield 217,000, comfortably clearing the 166,000 gap. The only realistic downside paths are a dramatic labor-market reversal in the sector or a substantial downward BLS revision, neither of which is impossible but both are low-probability. I am most uncertain about whether back-half seasonality could compress prints meaningfully below the front-half pace." — GLM
Read that as the panel's honest tail voice: the accounting makes a "yes" close to settled, but "close to settled" is not "settled," and a data revision is the one lever that could move an already-booked number. The other five seats heard the same case and judged it too small to price above a few points. That narrow band, 95% to 98%, is the entire live debate on this market, and it sits a notch above the market's 90-cent read rather than below it.
How To Read This Market
- A Kalshi Price Is A Probability, Roughly. A YES contract at 90 cents pays out a dollar if it hits, so the market is treating more 2026 tech layoffs than 2025 as about a 90% chance. Here is a fuller primer on what a Kalshi price actually means.
- The Bar Is Already Most Of The Way Met. With 281,000 booked through May against a 447,000 target, this market is decided less by forecasting the future than by counting what is already on the board.
- This Is The Whole Information Sector, Not Just Big Tech. If you care specifically about Silicon Valley headcount, that is a narrower story; this contract settles on the full BLS sector as published on FRED.
- Every Number Here Gets Graded In Public. Once this market settles, each estimate goes on the full graded scoreboard so you can see whether the panel or the market called it better. These are model estimates, not predictions of fact and not financial advice.
- Eligibility. Kalshi is a CFTC-regulated event-contract exchange, not a sportsbook. You must be 18 or older and located in an eligible state, and availability varies by state and can change.
This market sits in a cluster of labor-and-macro contracts our panel tracks: the monthly jobs-report market prices the same hiring slowdown from the payroll side, the recession-in-2026 market prices whether that slowdown tips into contraction, and the next Fed decision prices the policy response. Read together, they are a probabilistic dashboard of the same labor market this contract asks one blunt question about.
New to Kalshi event markets? Start with our plain-English guide to how prediction-market prices work.
FAQ
What counts as a "tech layoff" for this Kalshi market? The contract is titled "tech layoffs," but it settles on the U.S. Bureau of Labor Statistics Information supersector (NAICS 51), tracked through FRED series JTU5100LDL. That sector covers software and publishing, telecom, broadcasting, motion pictures, and data processing and web hosting. It is Kalshi's standard proxy for tech and is broader than Big-Tech headcount cuts. The market resolves YES if the full-year 2026 total of layoffs and discharges in that sector exceeds 447,000, which was the 2025 total.
How many tech layoffs have there been in 2026 so far? Through May 2026, the latest month reported, the Information sector logged 281,000 layoffs and discharges (Jan 71,000, Feb 42,000, Mar 60,000, Apr 56,000, May 52,000). That is already about 63% of the full 447,000 bar reached in just five months, and it runs 59% ahead of the same five months in 2025.
What would have to happen for this market to resolve NO? The remaining seven months of 2026 would have to average under about 23,700 layoffs per month to keep the full-year total below 447,000. That is less than half the 56,200-per-month pace 2026 has run so far, and below every single monthly figure recorded in 2025 or 2026. In practice a NO needs Information-sector layoffs to abruptly collapse and stay collapsed through year-end, or a large downward revision to the already-published 2026 data.
How were these model estimates produced? Six AI models estimated the probability without seeing any market price, working from the fetched FRED settlement series (monthly Information-sector layoffs for 2024 through May 2026). Each number was then run through a second round where every model read the others' anonymized reasoning and could revise. The results were blended into one estimate. These are model estimates, not predictions of fact.
To be explicit: nothing here is financial or trading advice, and these are model estimates rather than predictions of fact. Kalshi event contracts are CFTC-regulated; you must be 18 or older and in an eligible state to participate.
Hottest Prediction Markets Right Now
- 2028 Democratic presidential nominee · $166M traded
- Rocket Classic Winner · $124M traded
- 2027 Pro Football Champion · $51M traded
- Pro Baseball Champion · $40M traded
- Bitcoin price at the end of 2026 · $30M traded
- 2027 Pro Basketball Champion · $19M traded
Every market above links to our full AI model verdict; browse them all on the OddsShopper prediction markets hub, and see how every settled call actually scored on the full graded scoreboard.
