Global Engineering Intelligence Brief
The Great
Rebalancing.
Inside the 2026 software development economy — where AI adoption is universal but trust is falling, salaries diverge more than 4x across borders, and the talent map is shifting east. A data-led view for engineering & business leaders.
84%
of developers now use or plan to use AI tools in their workflow
41%
of all code written globally is now AI-generated
$6.31T
forecast worldwide IT spend in 2026, up 13.5% year over year
180M+
developers now active on GitHub, up 36M in a single year
01 AI Adoption & Trust
Everyone has adopted AI. Fewer people trust it.
Usage and confidence are moving in opposite directions — the defining paradox of software development in 2026.
Adoption vs. trust, 2024 → 2025
Trust AI accuracy2024
40%
Trust AI accuracy2025
29%
Positive sentiment2025 (was 70%+)
60%
What's driving the distrust
66%
of developers say debugging AI-generated code takes longer than writing it themselves.
46%
actively distrust AI output accuracy, versus 33% who trust it and just 3% who "highly trust" it.
52%
of developers avoid autonomous AI agents entirely or use only simple assistant modes.
"AI outputs are almost right — but not quite." A near-universal complaint, and the reason four out of five developers say they still verify before they ship.
STACK OVERFLOW DEVELOPER SURVEY, 2025
02 Languages & The GitHub Ecosystem
TypeScript dethrones Python and JavaScript.
The biggest shift in language rankings in over a decade — type safety is winning as AI writes more of the code.
Contributor growth, year over year (2025)
TypeScriptnow #1 language
+66.6%
PythonAI & data science
+48.8%
JavaScriptslowest of the three
+24.8%
180M+
developers now on GitHub, +36M added in 2025 alone — one every second
~80%
of new repos use just 6 languages: Python, JS, TS, Java, C++, C#
#1
India overtakes the U.S. in total open-source contributors for the first time
India added 5M+ new GitHub developers in 2025 (14% of global signups) and is projected to reach 57.5M developers by 2030 — one in three new global signups.
03 Global Compensation
A four-times pay gap for the same job title.
Median annual software engineer compensation by market, 2026 (USD).
SwitzerlandEurope's highest
$148.3K
United Statesnational median
$147.5K
Canadanational average
$117.0K
Western Europerange midpoint
~$96.5K
India (Bengaluru)leading tech hub
$40.4K
Eastern Europerange midpoint
~$35.4K
Senior AI/ML engineers in the U.S. earn $200K–$312K — versus $55K–$92K in Eastern Europe and $30K–$60K in India for the same seniority.
UVIK SOFTWARE, GEOR & ROBERT HALF 2026 COMPENSATION DATA
04 Hiring, Layoffs & The AI Divide
Two job markets, moving in opposite directions.
General software engineering roles are shrinking as AI/ML-specialist demand accelerates.
Job postings vs. Feb 2020 baseline
General SWE rolesvs. pre-pandemic
−49%
ML engineer rolesvs. pre-pandemic
+59%
AI skills in job ads2022 → 2026
42%
The 2026 snapshot
122.5K
global tech layoffs recorded in 2025, roughly 20% below the 2024 total.
67K
software engineering openings in Q1 2026 — the highest count since early 2023.
9%
of companies say AI has fully replaced roles; 45% say it only partially reduced hiring need.
Bay Area median time-to-hire for engineers stretched from 38 to 67 days within a single year, even as postings ticked back up — a more selective, slower-moving market on both sides.
05 Remote & Hybrid Work
The office pendulum swings back — partway.
Return-to-office mandates accelerated sharply in 2026, but flexibility hasn't disappeared.
87% fully on-siteQ2 2026 postings — up from 65% in Q4 2025
10% hybridsome in-office presence required
3% fully remoteno office presence required
−33%
lower voluntary turnover for hybrid teams vs. fully in-office, per peer-reviewed research
RETENTION WIN
65%
of Gen Z & Millennial employees would leave a job that eliminated remote flexibility
TALENT RISK
+23%
longer time-to-fill at S&P 500 firms after strict return-to-office mandates
HIRING COST
06 Cloud Infrastructure
The Big Three still control two-thirds of the cloud.
Global cloud infrastructure market share, early 2026 — even as AI reshapes growth rates.
AWS — ~30%global infrastructure leader
Azure — ~24%fastest growth among the Big Three at scale
Google Cloud — ~13%fastest % growth rate, +63% YoY
All others — ~33%Oracle, Alibaba, IBM & regional clouds
$129B
quarterly cloud infrastructure spend in Q1 2026, up 35% year over year
19%
of all cloud spending now goes to AI workloads, up from 8% in 2023
$1T
global cloud market approaching this threshold in 2026, from $918B
07 Supply Chain & Security
Open source is now a primary attack surface.
Malicious package volume accelerated sharply as attackers moved from typosquatting to trusted-maintainer takeovers.
+73%
surge in malicious open-source packages detected in 2025 alone.
1.23M+
cumulative malicious packages now cataloged across npm, PyPI, Maven, NuGet & Hugging Face.
99%+
of all open-source malware detections occur on the npm registry specifically.
65%
of organizations experienced a software supply chain attack in the past year.
581
mean vulnerabilities per codebase in 2026 — more than double the 280 recorded one year earlier.
80%
of AI-suggested software dependencies carry some form of security or licensing risk.
08 DevOps & Delivery Performance
AI is accelerating output — and exposing the seams.
DORA's 2025–2026 research finds individual throughput rising sharply while quality signals weaken underneath it.
Share of teams hitting the elite DORA benchmark
Elite deploy frequencymultiple times / day
16.7%
Lead time < 1 hourcommit to production
9.4%
Elite change-failure rate0–2% of deploys fail
8.5%
Per-developer gains attributed to AI assistance
+21%
more individual tasks completed with AI assistance
THROUGHPUT
+98%
more pull requests merged per developer with AI assistance
THROUGHPUT
+441%
increase in median pull-request review time year over year
QUALITY RISK
The trade-off: 31% more pull requests now merge with no review at all — meaningful throughput gains, paired with emerging quality and stability risk.
09 Talent, Diversity & Wellbeing
The workforce behind the workflow.
Representation gaps remain wide, and burnout is now a majority experience, not an edge case.
27–28%
of the global tech workforce is women, little changed over the past decade
~21%
of U.S. software developers specifically are women, per BLS workforce data
18%
of machine learning engineer roles globally are held by women
80%+
of developers report having experienced burnout at some point in their career
58%
moderate-to-extreme burnout in tech, third-highest of any industry measured
77%
of employees say AI tools have added to their workload, not reduced it
