The Layoffs Board
source: layoffs-view engine · as of 2026-07-13
How to read it, honestly: a cut only counts toward the AI figure when the company itself ties it to AI. Cyclical or unstated cuts (Intel, the semis) show in grey and never inflate the AI total. Rows dated 2024–25 are recency-flagged.
Amazon
The Read
Cutting 'in part to fund the hundreds of billions invested in AI'; some roles described by staff as AI-replaced.
Signal convergence: moderate
Insider-selling pressure (engine score)
25.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Intel
The Read
Massive foundry-turnaround restructuring — largely CYCLICAL, not AI-replacement. Flagged: does not smell like AI.
Signal convergence: moderate
Insider-selling pressure (engine score)
20.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Accenture
The Read
Reshaping consulting around GenAI — cutting roles automation absorbs while retraining 500k+. The one clean broad-market AI case. (2025-dated — recency flag.)
Oracle
The Read
Record results, record cuts — 21k cut while quarterly net income +27%; savings explicitly redirected to AI data centers.
Signal convergence: active
Insider-selling pressure (engine score)
45.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Dell
The Read
Workforce down ~10% (108k→97k) while projecting AI-optimized server revenue could double FY27 — cut the old base, bet the company on AI hardware.
Signal convergence: active
Insider-selling pressure (engine score)
60.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
IBM
The Read
~9k cut (3%) from back-office/HR roles it says AI (askHR) now performs; reinvesting into AI consulting headcount.
Signal convergence: watch
Insider-selling pressure (engine score)
30.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Meta
The Read
Framed as efficiency/offsetting AI investment; ~7k reassigned into new AI teams. Record ad revenue, record cuts.
Signal convergence: moderate
Insider-selling pressure (engine score)
25.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Alphabet
The Read
~5k across voluntary-exit programs + unit cuts alongside a ~$85B AI capex guide; stock +115% 1-yr — a clean 'winning and still cutting' quadrant point.
Signal convergence: moderate
Insider-selling pressure (engine score)
23.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Salesforce
The Read
Support & sales-dev roles cut/redeployed as Agentforce agents absorb the work; CEO cites AI doing a large share of tasks.
Signal convergence: moderate
Insider-selling pressure (engine score)
22.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Microsoft
The Read
States cuts are NOT direct AI replacement, but restructuring runs alongside tens of $B AI/DC capex; 4k+ redeployed. Our aggressive read, held pending contrary data.
Signal convergence: active
Insider-selling pressure (engine score)
30.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
PayPal
The Read
New 'AI transformation & simplification' team redesigning processes function by function; CEO: 'aggressively adopt AI.'
Cisco
The Read
Restructuring toward AI/security but largely cost-driven. AI-partial at best — flagged.
Signal convergence: watch
Insider-selling pressure (engine score)
44.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Broadcom
The Read
2024 VMware-integration cut — NOT 2026, NOT AI. Stale + cyclical; kept for L1 coverage, flagged.
Signal convergence: active
Insider-selling pressure (engine score)
57.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Intuit
The Read
Cut ~10% to redirect toward 'breakthrough AI experiences,' rehiring into AI roles — a reallocation, not a shrink.
Signal convergence: watch
Insider-selling pressure (engine score)
43.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Atlassian
The Read
Cut ~10% to rebalance toward AI; CEO: 'disingenuous to pretend AI doesn't change the number of roles required.'
Signal convergence: moderate
Insider-selling pressure (engine score)
48.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Cloudflare
The Read
Performance + AI-efficiency framing on the cut; reinvesting into AI/product.
Signal convergence: moderate
Insider-selling pressure (engine score)
50.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Qualcomm
The Read
Minor 2024 cut; stale, cyclical. L1 coverage cell, flagged.
Signal convergence: moderate
Insider-selling pressure (engine score)
47.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Apple
The Read
No broad cuts — L2 completeness cell; not a cut-to-fund-compute case.
Signal convergence: watch
Insider-selling pressure (engine score)
22.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
CoreWeave
The Read
Building, not cutting — the pure compute-supply bet. Contrast within L2.
Signal convergence: active
Insider-selling pressure (engine score)
50.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
ServiceNow
The Read
No confirmed 2026 layoff — an AI-platform winner (Now Assist), hiring not cutting. Corrected from pending: not a cutter.
Signal convergence: watch
Insider-selling pressure (engine score)
12.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Adobe
The Read
No confirmed 2026 cut; GenAI (Firefly) product pivot underway — a watch, not a cutter. Corrected from pending.
Signal convergence: active
Insider-selling pressure (engine score)
60.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
OpenAI
The Read
Zero cuts — hiring aggressively. The compute buyer the incumbents are cutting to pay. The contrast.
Signal convergence: active
Insider-selling pressure (engine score)
17.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Anthropic
The Read
Zero cuts — hiring. Shown as a hiring/zero-cut contrast; no claim about the company.
Signal convergence: active
Insider-selling pressure (engine score)
17.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
Palantir
The Read
Zero cuts — AI-native beneficiary, expanding.
Signal convergence: moderate
Insider-selling pressure (engine score)
68.0Engine score 0–100 — reads selling pressure from form-4 patterns. Pressure signal, not a dollar total; elevated score means more selling than the baseline, not intent.
xAI
The Read
Zero cuts — hiring/scaling.
Signal convergence: active
The aggregate view
The five charts that frame the headline: driver mix, monthly cadence, capex versus payroll freed, truth ladder, and the biggest decisions. Aggregate reads — the per-firm detail lives in the console above.
The macro backdrop
The build-out is up ~77% while the tracked cuts are up ~83%. Roughly 23% of 2026 cuts cite AI — the rest are cyclical or unstated, and we do not fold those into the AI total.
A handful of decisions drive the headline
The largest cuts by headcount, coloured by attribution. Grey bars are cyclical or not-AI and are never counted toward the AI-driven figure — Intel's cyclical cut is the one to watch here.
The driver mix
Every tracked cut sorted by why it happened. The centre number is the strict AI-driven total; the cyclical slice sits apart, not inside it.
Layoffs by month, 2026
2026-dated cuts only. AI-driven in red, cyclical / not-AI in grey, stacked by month with the peak flagged.
Capex versus the payroll it freed
AI capital spending against the payroll those cuts freed (headcount × ~$185k). Where the two track, that is cut-to-fund-compute; grey rows are cuts that are not AI-attributed.
The truth ladder
How solid each number is. Company-confirmed cuts — put on the record in a filing, earnings call, or release — held apart from credible press. Our sourcing discipline made visible; rumors sit in the radar, never in the totals.
Record results, record cuts — the quadrant
Stock one-year return against 2026 cuts, sized by AI capex, coloured by driver. The tell: winners still cutting. This lands when the trailing-return feed fills the X-axis.