2026 State of Tech Talent Report: Not a jobs crisis, but a skills crisis with an upskilling answer

Confidence 0.75 · last confirmed 2026-06-08

A 38-page annual survey report from Linux Foundation Research (authors Marco Gerosa of Northern Arizona University; Adrienn Lawson and Anna Hermansen of The Linux Foundation; foreword by Clyde Seepersad, SVP & GM Education). Based on an online survey fielded February 2026 with 400 participants worldwide responsible for hiring, training, and managing technical talent. It is the wiki’s first tech-talent-market survey anchor — a workforce-side complement to the firm-side and economy-side AI-adoption sources, and an empirical counterweight to the “AI is eating IT jobs” narrative.

TL;DR

• The headline thesis: “Not a jobs crisis, but a skills crisis with an upskilling answer.” AI is not reducing demand for technical talent — it is raising the bar for what that talent must do. The binding constraint is operationalising AI, not accessing it.
• AI is not eating all IT jobs: aggregated net hiring effect +26% (2025), +31% (2026); only the largest organisations report a negative net effect (−4%). 97% plan to implement AI; 55% expect significant value in software development.
• The skills gap is full-stack, not just AI: understaffing in AI (47%), cybersecurity (40%), cost optimization (36%), platform engineering (34%); AI security and AI operations each affect 57% of organisations. Most orgs lag on PARK-stack (foundational AI-infrastructure) implementation.
• Agentic AI raises serious security risks and most organisations are not ready; security concerns (48%) and budget constraints (47%) are the leading barriers to getting value from new technologies (security/privacy is now the top barrier).
• Upskilling is the answer: upskilling existing staff (57%) is the primary response to talent gaps, ahead of external hiring (49%); rated important by 94% for closing AI gaps and preserving institutional knowledge. Upskilling is favoured over hiring by 7.9× (business context), 7.7× (retention), 7.3× (team cohesion), 5× (lower total cost); orgs are 3.5× more likely to upskill than hire across strategic domains.
• Talent signals are shifting: technical training (93%) ranks above compensation (91%) as a retention strategy; 76% of hiring managers consider certifications important; degrees/titles/years-of-experience are “proving insufficient” as roles change faster than those signals.

What was actually ingested

Full 38-page report (executive summary, four findings sections, conclusion, methodology + demographics, appendix; converted via pdftotext). Identity verified against the cover and citation block. Survey: online, Feb 2026, n=400 worldwide. The report’s own external anchors include McKinsey State of AI 2025, BCG Where’s the Value in AI? Only 5% Are Getting It at Scale, Deloitte State of GenAI in the Enterprise, Bain Widening Talent Gap, BCG AI Transformation Is a Workforce Transformation, and LinkedIn skills-based-hiring trends.

Key findings, with detail

”Not a jobs crisis, but a skills crisis”

The foreword (Seepersad) states the thesis bluntly: “AI does not appear to be reducing demand for technical talent. It is raising expectations for what that talent must be able to do … The issue is not simply a lack of people. It is a lack of capability — specifically, the ability to apply skills across systems, workflows, and real-world environments.” This is the workforce-side mirror of the firm-side micro-productivity-trap finding (the constraint is operationalisation, not technology) and of BCG’s 5%-get-value-at-scale anchor — the report cites both directly.

AI is not eating all IT jobs (net hiring is positive)

The +26%/+31% aggregated net hiring effect, with only the largest orgs negative (−4%), is the survey-side complement to Brynjolfsson, Chandar & Chen’s ADP-payroll evidence. The two are not identical: Brynjolfsson et al. find a compositional twist (entry-level decline in the most AI-exposed occupations) that the Global report’s aggregate doesn’t surface — but the Europe sibling report does (see Europe report, −3% entry-level). Together they qualify the headline: aggregate IT demand rises and the junior tier is where the pressure concentrates.

Upskilling over hiring — and why

The report’s distinctive contribution is the quantified upskilling-vs-hiring preference (7.9×/7.7×/7.3×/5× across business context, retention, team cohesion, cost) and the framing of upskilling as institutional-knowledge preservation, not just gap-closing (“upskilled talent doesn’t take wing and fly, they grow roots”). This is the strongest workforce-side evidence in the wiki for the durable-skills thesis and complements Forsgren & Macvean on durable core skills.

Skills-based signals over credentials-of-record

Certifications (important to 76% of hiring managers) and demonstrated, job-relevant skills are displacing degrees/titles/tenure as trust signals — a shift with an obvious interest-alignment caveat (the publisher sells certifications; see source-quality flag).

Dynamic-capabilities reading (Warner & Wäger)

• digital-transforming/improving-digital-maturity (core) — the entire report is about “identifying digital workforce maturity,” “external recruiting of digital natives,” and “leveraging digital knowledge inside the firm” (upskilling + institutional-knowledge preservation).
• digital-transforming/redesigning-internal-structures — organisations “redefining roles in real time and expanding the scope of responsibilities” across software, infrastructure, data, and security.
• contextual/internal-barriers — skills/capability gaps, security concerns, and budget constraints named as the barriers to getting value from AI.

(Roles inherited from cell defaults: chro, cdo, cio, ceo.)

Linked entities and concepts

Existing pages (touched or referenced):

• ai-employment-effects — heavy: net-hiring evidence; “AI not eating IT jobs”; role redefinition.
• durable-skills — heavy: upskilling, institutional knowledge, skills-over-credentials.
• enterprise-ai-adoption — moderate: 97% AI adoption, value-at-scale gap, security as the binding barrier.
• automation-vs-augmentation — light: roles redefined/expanded (augmentation), not eliminated.
• responsible-ai — light: agentic-AI security risks and readiness gap.

New pages created with this ingest:

• The Linux Foundation — entity (organization): promoted on cross-source presence (this report + the Europe report + the Headroom talk channel).
• Marco Gerosa — entity (person): lead author of both 2026 State-of-Tech-Talent reports.
• Adrienn Lawson — entity (person): co-author of both reports (Linux Foundation Research).

Dangling (single-source mention, deferred): Anna Hermansen (Global-report co-author only), Clyde Seepersad (foreword).

Source-quality flag

• Strengths: a real survey (n=400, methodology + demographics disclosed, Data.World access); cross-referenced against independent anchors (McKinsey, BCG, Deloitte, Bain); the wiki’s first workforce-side empirical anchor on the AI-and-jobs question; directly corroborates the Brynjolfsson-canaries reading at the aggregate level.
• Caveats: interest alignment — Linux Foundation Education sells training and certifications, so the report’s central prescription (upskilling + certifications) aligns with the publisher’s commercial interest; self-reported survey (intentions like “97% plan to implement AI” overstate realised deployment); n=400 global is modest for fine-grained breakdowns. Per CLAUDE.md, a vendor-interested survey is held at ≤0.75 unless independent sources agree — several do (Brynjolfsson, BCG, Bain), supporting 0.75.
• Confidence: 0.75.

Open questions for the wiki

• The PARK stack (foundational AI infrastructure; Ray/Distributed Computing the lowest-deployed component) is named but not defined here — worth pinning down if a second source uses it.
• Aggregate-positive but entry-level-negative: the reconciliation with Brynjolfsson lives in the Europe report’s −3% entry-level finding — track whether future LF reports surface the entry-level cut globally.