SASIGNAL ATLASCross-industry intelligence / Research desk
SIGNAL ATLAS / RESEARCH DESK

Near-universal AI coding tool adoption with falling developer trust

Signal report · Enterprise software & SaaS · Original Phase 1 research

Near-universal AI coding tool adoption with falling developer trust

Enterprise software & SaaS · Sector 02 (rank 5 of 25) · T-02-06
Score 80.8/100 · Evidence factor 1.00 · Confidence high · triangulated · Last verified 2026-09-15

In one sentence. AI coding assistance is now the default in professional software development, but confidence in it is going the other way.

Why it matters

Adoption is settled; value capture is not. The DORA finding that AI amplifies existing organisational strengths and weaknesses means the same tool produces opposite results at different companies, which undermines any blanket productivity assumption in headcount or budget planning. The trust gap is also the mechanism that keeps humans in the loop - and therefore keeps some seats in the seat count.

What is happening

AI coding assistance is now the default in professional software development, but confidence in it is going the other way. Stack Overflow found 84% of developers using or planning to use AI tools, up from 70% in 2023, while trust fell to 29% from roughly 40% in 2023 - an 11-point drop in a single year. Google's DORA 2026.01 research, based on nearly 5,000 practitioners and over 100 hours of interviews, concluded that returns come from the surrounding organisational system rather than the tools, and documented an instability tax where change-failure rates rise as code volume increases faster than delivery practices can absorb.

First observable signal. Stack Overflow's 2023 survey already showed 70% adoption; the inflection was the trust reversal visible in the 2025 survey wave reported 2026-02-18.

Evidence

Claim Type Date Source
84% of developers reported using or planning to use AI tools, up from 70% in 2023; trust fell to 29%, down 11 percentage points from 2024 and from roughly 40% in 2023. fact 2026-02-18 Stack Overflow (A)
DORA 2026.01 surveyed nearly 5,000 technology professionals with over 100 hours of qualitative interviews and found the greatest returns on AI investment come not from the tools themselve… fact 2026-05-11 InfoQ, reporting Google Cloud DORA research (B)
DORA documented an instability tax - an illustrative rise in change failure rate from 5% to 6% following adoption - and a J-curve in which productivity dips before gains materialise. estimate 2026-05-11 InfoQ, reporting Google Cloud DORA research (B)
DORA's illustrative model shows approximately 39% first-year ROI with an eight-month payback for a 500-person engineering organisation. estimate 2026-05-11 InfoQ, reporting Google Cloud DORA research (B)
Cursor reported over 1 million paying customers, 2 million-plus users, approximately 50,000 enterprise teams and representation in nearly 70% of the Fortune 1,000. marketing 2026-04-18 The Next Web (Ana Maria Constantin) (B)

Companies and products

Companies. Anysphere (Cursor, now SpaceXAI), Microsoft/GitHub, Anthropic, Google, Atlassian, JetBrains, Sourcegraph
Products. Cursor, GitHub Copilot, Claude Code, Gemini Code Assist, JetBrains AI Assistant
Funding. Anysphere raised $2B at a $50B valuation in April 2026 (a16z and Thrive Capital lead, Nvidia strategic), after a $29.3B valuation in November 2025; acquired by SpaceX for $60B, closed 2026-08-14.

Impact

Industry. Beneficiaries: Model providers and dev-tool vendors, Code review, testing and observability vendors, Senior engineers whose review capacity becomes the constraint. Losers: Junior/entry-level developer hiring, Stack Overflow and Q&A communities, Low-differentiation outsourced coding shops.
Consumer. Documented at the practitioner level by two independent instruments (Stack Overflow n undisclosed; DORA n~5,000). Vendor-side counts (Cursor's 1M paying customers) are self-reported and labelled marketing.
Regulatory. Low today. Emerging exposure: EU AI Act transparency obligations (effective 2026-08-02) and open questions on copyright provenance of generated code.
Geography. US, EU, IN, CN, GLOBAL

Risks and counter-forces

  • Adoption without measurable throughput gain
  • Rising change failure rates and technical debt
  • Trust decline could stall enterprise standardisation

Counter-trends. Falling trust caps autonomous use, Code review and security remain human bottlenecks, DORA's instability finding argues against naive headcount reduction

Score breakdown

Pillar Score Dimensions
Momentum (30%) 90 velocity 4 · adoption 5 · capital 5 · revenue 4
Reach (25%) 95 breadth 4 · depth 5 · geography 5 · demand 5
Durability (25%) 80 persistence 3 · maturity 4 · strategic 5
Consequence (20%) 50 regulatory 1 · social 4
Evidence 100 quality 5 · diversity 5 → ceiling 100

Stage mainstream · Direction accelerating · Horizon Immediate (0–12 months) · Reading: Act on it

What to watch next

  • T-02-07 Software hiring contraction: postings a quarter below pre-pandemic levels
  • T-02-11 Strategic acquirers re-rating developer tools: the $60B Anysphere precedent
  • T-02-12 Build-versus-buy shifting toward build as coding agents cut internal software cost
  • T-02-20 Autonomous agentic AI running production enterprise workflows

Sources

  1. Mind the gap: Closing the AI trust gap for developers — Stack Overflow, 2026-02-18. Tier A. https://stackoverflow.blog/2026/02/18/closing-the-developer-ai-trust-gap/
  2. New DORA Report Claims Strong Engineering Foundations Drive AI Return on Investment (DORA 2026.01) — InfoQ, reporting Google Cloud DORA research, 2026-05-11. Tier B. https://www.infoq.com/news/2026/05/dora-roi-ai-assisted-dev-report/
  3. Cursor in talks to raise $2B at $50B valuation after hitting $2B ARR in three years — The Next Web (Ana Maria Constantin), 2026-04-18. Tier B. https://thenextweb.com/news/cursor-anysphere-2-billion-funding-50-billion-valuation-ai-coding
  4. SpaceX completes record $60 billion acquisition of AI coding platform Cursor — Yahoo Finance, 2026-08-14. Tier B. https://finance.yahoo.com/technology/ai/articles/spacex-completes-record-60-billion-131311785.html

Generated from record T-02-06 via the canonical article template (18-editorial-formats.md). Research date 2026-09-15. Scores per 14-scoring-framework.md. Forecasts are conditional, never certainties.

Research provenance
Source artifact
06-sample-reports/02-enterprise-software.md
Corpus date
15 September 2026
Prepared for this site
16 September 2026
Site publication
18 September 2026
Verification
Inherited; not fully rechecked