← Back to scans

More tech layoffs in 2026 than in 2025? — Yes

KXLAYOFFSYINFO-26-494000 · Economics · 2026-05-04
82%
Agent
87%
Market Price
-4.9%
Edge
50%
Confidence
Volume: 31,298,484
Spread: 0.1c
Days to resolution: 301
Markets in event: 1
Final Rationale
My independent estimate of 67% inconsistently applied tracker YTD data (92-114k by May) as a strong downward signal while simultaneously acknowledging that BLS NAICS 51 resolution methodology captures a much broader category that trackers undercount — I ACCEPT this critique and remove the inconsistent downward pressure. I also ACCEPT the back-loaded pipeline argument: Meta (8k), Microsoft (8.75k), Oracle (30k), Amazon (30k continuation) are announced or in-progress for 2026, with 8 months of resolution time remaining and structural AI capex-vs-headcount dynamics intact. I PARTIALLY ACCEPT the regime-change argument — the historical base rate of 'peaks moderate' is weakened but not nullified. The high-volume market at 88.9% with sustained upward trending over 67 days reflects informed updating I can't fully dismiss, but I retain ~7pp of divergence due to genuine threshold ambiguity (447k description vs 494k ticker) and the possibility that H2 2026 sees fewer new announcements than the pipeline suggests.
Price vs. prediction
How the market moved after the agent's call. Solid blue line is the actual market price; dashed line is the agent's static forecast at $t{=}0$.
$n = 58$ follow-ups
Re-scan Context
This market has been scanned before. Previous predictions:
DatePredictedMarket PriceConfidence
2026-04-03 92% 86% 80%
1. Planner Opus · Fermi decomposition + tool plan
Tools chosen: kalshi_data article_search web_search fred_data code_execution
Sub-questions (Fermi decomposition)
  1. Will the pace of tech/information sector layoffs in early 2026 (Q1-Q2) be running ahead of the 2025 pace on an annualized basis? (w=0.35)
  2. Will macroeconomic conditions (recession risk, AI-driven restructuring, or major tech downturns) in 2026 sustain or accelerate layoffs beyond 2025 levels? (w=0.35)
  3. Has the information sector already accumulated more than ~200,000 layoffs through the first 4 months of 2026 (putting it on track to exceed 447,000 by year-end)? (w=0.30)
Planner reasoning
This question resolves YES if 2026 information sector layoffs exceed 447,000 (the stated 2025 level). The market ticker specifically references 494,000, suggesting the contract may be priced around a higher threshold — this distinction matters. Key analytical framework: 1. **Base rate of year-over-year escalation**: Tech layoff cycles have shown clustering — 2022-2023 saw massive waves (~260k in 2023), 2024 moderated somewhat, 2025 reportedly hit ~447k. The question is whether 2026 continues an upward trend or reverts. 2. **Structural drivers**: AI-driven restructuring (replacing workers), post-bubble correction in tech hiring, economic slowdown, and rising interest rates are persistent headwinds. These suggest elevated layoffs could continue. 3. **Temporal analysis**: With 301 days until resolution (March 2027), and today being May 4, 2026, roughly 4 months of 2026 data exists. If layoff trackers (Layoffs.fyi, Challenger Gray & Christmas, BLS JOLTS) show ~150,000+ information sector layoffs already in Jan-April 2026, that would be on pace to exceed 447k. The remaining ~8 months (May-December) would need to contribute ~300k+ if early pace is ~150k. 4. **Critical distinction**: The market threshold is 494,000 (not 447,000), meaning the market is asking about exceeding a higher bar than the minimum resolution threshold. This suggests some embedded difficulty. 5. **Causal mechanism check**: For YES, you need: continued AI-driven job displacement, possible macro deterioration (tariffs, slowdown in 2026), and large tech companies (Meta, Google, Amazon, Microsoft) continuing restructuring programs. 6. **Reference class**: In years with large tech corrections or macroeconomic stress, information sector layoffs can spike dramatically. The 2023 wave was ~260k (Challenger data). If 2025 was truly ~447k, 2026 exceeding that is a high bar historically.
2. Research Parallel tool calls · legacy v2
ToolStatusTimeItemsSummary
kalshi_data OK 2.1s -
article_search OK 0.7s -
web_search OK 53.5s -
fred_data OK 1.3s -
code_execution OK 0.0s -
3. Evidence Synthesis Sonnet · 19 items · legacy v2
#StrengthCredibilityDirectionSourceClaimPriced In
1 STRONG 72 ↓ DOWN web_search As of early May 2026, Layoffs.fyi reports 92,000+ tech layoffs YTD; SkillSyncer tracker reports 113,863 workers impacted in 179 events; TrueUp reports 95,878 impacted — all figures significantly below the ~150,633 needed to be on pace for 447k by year-end. Yes
2 STRONG 82 ↑ UP article_search January 2026 saw 108,435 total U.S. layoffs (Challenger data) — highest January since 2009 — with technology sector accounting for 22,291 of those cuts, up 118% YoY. Yes
3 STRONG 88 ↑ UP article_search Meta announced 8,000 job cuts (~10% of workforce) in late April 2026 with layoffs scheduled May 20, plus 6,000 open positions closed; Microsoft announced voluntary buyouts affecting ~8,750 U.S. employees with notifications going out May 7. Yes
4 STRONG 80 ↑ UP web_search Oracle cut 30,000 employees and Amazon cut ~30,000 jobs since October 2025, representing major early-2026 tech layoff contributors not yet fully reflected in some trackers. Yes
5 MODERATE 65 ↓ DOWN code_execution To be on pace for 447,000 layoffs by year-end, approximately 150,633 layoffs were needed by May 4, 2026; current tracker data (92,000-113,863) falls 37,000-58,000 short of this threshold depending on source. No
6 MODERATE 75 ↑ UP web_search Major upcoming layoff waves not yet counted: Meta's 8,000 cuts execute May 20; Microsoft's ~8,750 buyout decisions close around June 6; additional Meta layoffs planned for H2 2026 — suggesting significant near-term additions to YTD totals. No
7 MODERATE 60 NEUTRAL article_search There is a critical methodological gap: tracker data (Layoffs.fyi, TrueUp, SkillSyncer) captures 'tech company' layoffs, not specifically the BLS 'information sector' definition used for contract resolution — the BLS figure may differ substantially. No
8 STRONG 85 ↑ UP article_search 2025 total U.S. layoff announcements reached 1.17 million through November — 54% higher than same period in 2024 and the highest since the 2020 pandemic — driven by AI restructuring, tariffs, and corporate rightsizing. Yes
9 STRONG 80 ↑ UP article_search Major tech companies (Meta, Microsoft, Amazon, Alphabet) are collectively spending ~$700 billion on AI infrastructure in 2026 while simultaneously cutting headcount, indicating AI-driven restructuring is a sustained structural driver. Yes
10 MODERATE 70 ↑ UP article_search Economists and industry experts quoted in late April 2026 warn an 'AI-driven labor crisis may be upon us today — not coming sometime in the future,' with AI replacing roles in content creation, customer support, data entry, and basic coding. Yes
11 MODERATE 92 ↓ DOWN fred_data Initial jobless claims as of April 25, 2026 were 189,000 — down 41,000 YoY — suggesting the broader labor market remains relatively tight despite tech-specific layoffs, which could constrain escalation. Yes
12 MODERATE 92 ↓ DOWN fred_data U.S. unemployment rate was 4.3% as of March 2026, up only 0.1 percentage points YoY — no macro recession signal yet, though tech-specific stress is elevated. Yes
13 MODERATE 65 ↑ UP article_search Geopolitical risk (Iran war escalation threat as of early April 2026) and tariff uncertainty are cited as macro headwinds for tech, adding recession risk that could accelerate layoffs beyond AI-driven restructuring alone. No
14 MODERATE 68 ↓ DOWN article_search Amazon's layoffs are described as 'not a direct result of AI' but reflect broader rightsizing from pandemic overhiring, suggesting some layoff drivers are temporary and may taper as rightsizing completes. Yes
15 STRONG 72 ↓ DOWN web_search As of May 3-4, 2026, Layoffs.fyi shows 92,000+ tech layoffs YTD; SkillSyncer shows 113,863; TrueUp shows 95,878 — all well below the ~200,000 threshold that would indicate the sector is on pace to exceed 447k by year-end. Yes
16 MODERATE 78 ↑ UP web_search Significant layoffs are announced but not yet executed: Meta's 8,000 cuts (May 20), Microsoft's ~8,750 buyouts (decisions by June 6), and additional Meta H2 2026 rounds could add 20,000-30,000+ to YTD figures in coming weeks. No
17 WEAK 45 ↑ UP article_search The Challenger data methodology (announced cuts) vs. BLS JOLTS/information sector definition may produce very different totals; the resolution metric (BLS information sector) is likely broader and could capture more workers than tech-tracker data. No
18 STRONG 80 ↑ UP kalshi_data The Kalshi market for this contract (threshold 494,000) is priced at 88.90% YES as of early May 2026, with a 30-day increase of +2.80% and very high volume (467,106 contracts/day average), suggesting the market strongly expects the threshold to be exceeded. Yes
19 MODERATE 55 NEUTRAL article_search The question description states 2025 information sector layoffs were ~447,000, yet the Kalshi contract resolves at 494,000 — a ~10.5% higher bar — creating ambiguity about which threshold actually governs resolution. No
Information Gaps
  • No BLS JOLTS or official information sector (NAICS 51) layoff data for Q1 2026 — all available data is from private trackers (Layoffs.fyi, TrueUp, SkillSyncer) using broader 'tech company' definitions that may not match the BLS information sector classification used for resolution
  • No monthly breakdown of 2025 information sector layoffs by quarter to establish a comparable pace baseline for YoY comparison
  • No confirmed total for Oracle's 30,000 cuts as to timing and whether they qualify as 'information sector' under BLS definitions
  • No data on Q2 2026 (April-June) Challenger Gray & Christmas sector-specific breakdown beyond January figures
  • No clarity on whether the resolution metric is BLS JOLTS information sector separations (which would be much larger and include voluntary quits) or only announced layoffs/involuntary separations
  • No data on whether announced layoffs (Meta May 20, Microsoft June) will be counted in 2026 or partially deferred
  • No information on H2 2026 planned restructuring programs at Alphabet, Apple, or other large information-sector employers
Key Uncertainties
  • Resolution methodology: BLS 'information sector' data (which includes telecommunications, publishing, broadcasting, and software) may differ dramatically from tech-tracker data — the sector is much broader than 'tech companies' and could already be running much higher
  • Whether the contract resolves at 447,000 (question description) or 494,000 (Kalshi ticker) — this ~10% gap significantly changes the probability assessment
  • Scale of H2 2026 layoff waves from companies with announced but unexecuted programs (Meta second round, potential Google/Alphabet restructuring, downstream UPS/logistics effects)
  • Macro tail risks: Iran conflict escalation, tariff-driven recession, or equity market correction could dramatically accelerate tech layoffs beyond current pace
  • AI displacement acceleration: whether AI productivity gains translate to faster-than-expected headcount reductions at large tech firms in H2 2026
  • Whether the 2025 baseline of 447,000 is itself accurate — if it was revised upward, the 2026 bar is higher
Market Context
Price: 0.889 · 7d: +0.04 · 30d: +0.028
Volume: VERY_HIGH — 467,106 contracts/day average is extremely high for a labor market prediction market · Depth: HIGH — 67 days of data with price range 66%-91.5% showing substantial trading history
Reliability: HIGH — very high volume suggests informed participants with significant capital at stake; the price movement from 66% floor to current 88.9% indicates substantial updating on new information over the contract lifetime; however, the discrepancy between the question description threshold (447k) and the ticker threshold (494k) introduces contract interpretation uncertainty
4. Independent Forecaster Opus · No market price · legacy v2 Did NOT see market price
Independent estimate: 67% (47% confidence)
base rate
62%
40% conf
evidence driven
78%
55% conf
contrarian
62%
45% conf
Spread
16pp
High disagreement
sq1: Will Q1-Q2 2026 pace exceed 2025 annualized?
55%
Base rate for layoff cycles continuing year-over-year escalation after a peak year is moderate (~40%). YTD trackers (92k-114k by May 4) are running BELOW the 150k pace needed for 447k extrapolation — pushes DOWN by ~15%. However, Meta (8k), Microsoft (8.75k), and pending H2 rounds will add meaningfully soon — pushes UP by ~10%. January was highest since 2009 with tech up 118% YoY — pushes UP by ~10%. Net: tracker data slightly underpaces but methodology gap with BLS information sector (which is broader: telecom, publishing, software) likely means actual BLS info sector numbers are higher. Estimate 55%.
sq2: Will macro/AI conditions sustain elevated layoffs in 2026?
75%
Base rate for sustained layoff environment when prior year hit highest since pandemic is high (~60%). AI-driven restructuring is a structural, not cyclical driver — $700B AI capex with simultaneous headcount cuts pushes UP by ~15%. 2025 announcements were 54% above 2024 indicating accelerating trend — pushes UP by ~10%. Counter: jobless claims at 189k (down 41k YoY) and unemployment only 4.3% suggest broader labor market is healthy, which could limit panic-driven cuts — pushes DOWN by ~10%. Net: structural AI displacement plus tariff/geopolitical risk supports continued elevated layoffs. 75%.
sq3: Has info sector already accumulated >200k layoffs in first 4 months of 2026?
45%
Tech trackers show only 92k-114k YTD as of May 4 — substantially below 200k. Pushes DOWN strongly by ~25%. However, BLS 'information sector' (NAICS 51) is a much broader category than what trackers capture — includes telecom (Verizon, AT&T cuts), publishing, broadcasting, data processing. BLS JOLTS layoffs/discharges in information sector typically run 30-50k/month even in normal times, so 4 months could easily be 150-200k+ on JOLTS basis. The methodology question dominates. If resolution uses Challenger announced layoffs in tech: probably not on pace. If using BLS JOLTS information sector separations: very likely on pace. Estimate 45% reflecting genuine ambiguity.
Show full reasoning trace
base rate: Year-over-year escalation in layoff cycles after a peak year is uncommon — typically peaks are followed by moderation. However, 2023 (~260k) → 2024 (moderation) → 2025 (~447k) shows non-monotonic pattern. Base rate for exceeding prior year peak: ~35-45%.
evidence updates: Strongest evidence DOWN: tracker data through May 4 (92-114k) is ~25-40% below pace needed. Strongest evidence UP: AI-driven structural restructuring with $700B capex, January 2026 being highest since 2009, multiple major announced waves (Meta, Microsoft, Oracle 30k, Amazon 30k) cementing structurally elevated layoff environment. The BLS information sector methodology likely captures more than tech trackers.
combination method: Weighted average across pace, structural drivers, and current accumulation. Sq2 (structural drivers) is most predictive for full-year outcome since 8 months remain.
final: Probability around 0.60-0.65. Key uncertainty: resolution methodology (BLS vs Challenger vs trackers) and whether threshold is 447k or 494k. Given question description specifies 447k (the lower bar), and structural AI/restructuring drivers remain potent, I estimate 0.62 with medium-low confidence due to methodology ambiguity.
ensemble: {'ensemble_count': 3, 'perspectives': {'base_rate': {'p_yes': 0.62, 'confidence': 0.4}, 'evidence_driven': {'p_yes': 0.78, 'confidence': 0.55}, 'contrarian': {'p_yes': 0.62, 'confidence': 0.45}}, 'spread': 0.16, 'individual_p_yes': {'base_rate': 0.62, 'evidence_driven': 0.78, 'contrarian': 0.62}}
5. Devil's Advocate Sonnet · legacy v2 Market price revealed
Asymmetric Evidence Reasoning Flaw Missing Info
Challenges
  1. The forecaster identifies a critical methodology question (BLS information sector NAICS 51 vs Challenger tech tracker vs Layoffs.fyi) but doesn't resolve it before forecasting. If resolution uses BLS information sector data — which the question description references ('2025 information sector layoffs were ~447,000') — then tracker YTD data of 92-114k is largely IRRELEVANT as a downward signal because trackers undercount the broader NAICS 51 category by a large factor. The forecaster acknowledges this in sq3 but then applies tracker pace as a 'STRONG DOWN' signal in sq1 anyway, which is inconsistent.
  2. The base rate framing is questionable. The forecaster says 'YoY escalation after a peak year is uncommon' implying ~35-45%, but 2025 was itself a YoY escalation over 2024, and the structural AI driver is described as new. Anchoring to a historical base rate that doesn't reflect the structural regime change may be inappropriate.
  3. The high-volume market at 88.9% with sustained upward trend (66%→89% over 67 days) reflects informed traders updating on the same evidence the forecaster sees. The forecaster doesn't articulate what specific information they have that the market is missing — only methodology ambiguity, which cuts both ways. With 301 days to resolution and a market that's directionally trending up, the forecast at 67% requires justification that isn't fully provided.
  4. Sq3's 45% answer for '>200k accumulated by April' doesn't cleanly feed into the year-over-year question. Even if Q1-Q2 underpaces 2025, H2 2026 announced waves (Meta, Microsoft, Oracle 30k, Amazon 30k continuation) could push the full year over. The combination method underweights the back-loaded nature of the announced 2026 layoff pipeline.
  5. Confidence of 0.467 seems appropriately humble given methodology ambiguity, but the point estimate doesn't reflect the asymmetric payoff: under most reasonable resolution methodologies (BLS NAICS 51 with ~447k baseline), the strong structural drivers favor YES more heavily than 67% suggests.
Suggested adjustment: +8pp
Adjust upward modestly because: (1) the forecaster inconsistently applies tracker YTD data as strong downward evidence in sq1 while acknowledging in sq3 that trackers don't match the BLS resolution methodology — this is a reasoning inconsistency, not a market-conformity argument; (2) the back-loaded 2026 layoff announcement pipeline (Meta, Microsoft, continuing Oracle/Amazon) is underweighted in a question with 8 months remaining; (3) the structural AI driver is treated as a normal evidence update rather than a regime change that invalidates the historical base rate. A forecast around 0.74-0.77 better reflects the internal logic of sq2 (75%) combined with the genuinely ambiguous sq1/sq3.
Market comparison: Forecaster: 67.3%, Market: 88.9%. The 21.6pp divergence is substantial. The market has high volume and has trended upward steadily over 67 days. Some divergence is justified by genuine resolution methodology ambiguity (447k vs 494k threshold; BLS vs Challenger), which the market may underprice. However, part of the gap reflects the forecaster's inconsistent treatment of tracker data and underweighting of structural AI drivers. A modest upward adjustment is warranted on reasoning grounds, not full conformity.
6. Aggregator Opus · final calibrated forecast
82% (50% confidence)
My independent estimate of 67% inconsistently applied tracker YTD data (92-114k by May) as a strong downward signal while simultaneously acknowledging that BLS NAICS 51 resolution methodology captures a much broader category that trackers undercount — I ACCEPT this critique and remove the inconsistent downward pressure. I also ACCEPT the back-loaded pipeline argument: Meta (8k), Microsoft (8.75k), Oracle (30k), Amazon (30k continuation) are announced or in-progress for 2026, with 8 months of resolution time remaining and structural AI capex-vs-headcount dynamics intact. I PARTIALLY ACCEPT the regime-change argument — the historical base rate of 'peaks moderate' is weakened but not nullified. The high-volume market at 88.9% with sustained upward trending over 67 days reflects informed updating I can't fully dismiss, but I retain ~7pp of divergence due to genuine threshold ambiguity (447k description vs 494k ticker) and the possibility that H2 2026 sees fewer new announcements than the pipeline suggests.