This page compiles developer-adoption data for AI coding tools, drawing on the Stack Overflow 2025 Developer Survey (n=49,000+), JetBrains’ January 2026 AI Pulse survey (10,000+ developers), GitHub Octoverse 2025, Google’s DORA reports, the DX engineering-impact report, and METR’s randomized controlled trial.
The AI Behavior Index is the research arm of OneChat AI, an integrated multi-model AI platform. We compile and analyze data from primary research sources to make AI adoption and market trends more accessible to journalists, researchers, and decision-makers.
Why adoption figures range from 84% to 91%: they measure different populations and questions. Stack Overflow’s 84% is “use or plan to use” (so active use is lower); JetBrains’ 90% is “regularly use ≥1 tool at work”; DX’s 91% comes from organizations already on its engineering-intelligence platform (a more AI-forward sample). Tool-share figures also differ between Stack Overflow and JetBrains because of timing (Claude Code launched into the window) and sampling — the spread is the story, not any single number.
Methodology notes: survey figures come from large multi-country developer surveys; tool-share is work-adoption (a developer can use several, so shares don’t sum to 100%). The METR result is a randomized controlled trial on experienced open-source developers and may not generalize to all developers or tasks. Vendor studies (GitHub, Cursor) report larger productivity gains (50–100%) than independent ones — we lead with independent figures and flag vendor-sourced ones. The use-vs-trust gap is consistent across Stack Overflow and DORA.
Sources used on this page:
- Uvik / Stack Overflow — AI Coding Assistant Statistics 2026 (adoption; trust; METR)
- Konabayev / JetBrains — AI Code Assistant Statistics 2026 (tool shares; daily use)
- Digital Applied — AI Coding Adoption: 50 Statistics 2026 (cross-source analysis)
- Panto / DX & DORA — AI Coding Statistics 2026 (productivity; defect rates)
Corrections or suggestions: research@aibehaviorindex.org