AI was expected to make software work faster, better, and more fulfilling. The Q3 2026 Dev Barometer tests those expectations against a full year of lived experience. One year in, only 21% of developers now spend most of their week writing new code.
Developers surveyed
705Countries represented
60Conducted by
BairesDevPublished on
Q3 2026Key Findings
What developers told us
Productivity Dividend
13 hrs
AI is now saving developers nearly two full workdays every week. But almost none of that time is going back to the developer. It's being reinvested into learning, reviewing, and debugging AI-generated work.
Deep Integration
42%
In twelve months, AI has moved from the edges of the developer workflow to inside it. 72% of developers say AI now assists with at least a quarter of their code, more than double a year ago.
The AI Pay Gap
5 vs 3
20% of developers say AI fluency helped them earn more in the past year. 12% saw AI cited as a reason to limit their raise, reduce headcount, or restructure their role. 61% report no direct impact at all.
The Human Skills Premium
15%
Among developers whose pay rose because of AI, AI tool fluency and prompting (29%) and system design/architecture (20%) led the way, but human skills outranked other AI expertises, including AI security (9%), AI quality assurance (6%), and data engineering (4%).
The Productivity Dividend
AI is delivering measurable time savings to developers. Those hours are going back into managing a more complex layer of work overseeing AI output.
What developers gained
“13 hours a week saved on coding with AI, up from 7 a year ago”
Developers are getting back nearly two working days a week. The share who say AI writes at least half their code jumped from 12% to 42% in a year.
Where those hours are going
Time reviewing AI-generated code has increased:
67%
Time debugging AI-introduced problems:
52%
Time correcting juniors over-relying on AI:
44%
We've spent a year asking developers, quarter by quarter, how AI is actually changing their work. What's emerged is a picture that shifts faster than any annual report could capture. Developers pointed to AI tool fluency, system design, and human skills as the top drivers of their raises. In a market moving this quickly, staying close to the people building with these tools is one of the best ways to see what's actually changing.
Rodrigo Outumuro
VP of People Experience
BairesDev
How AI restructured the role of the developer in a year
The productivity dividend is real. Developers are using the time to catch what AI produces, learn what AI can do, and carry the responsibility for the code AI writes on their behalf. That work is more demanding than it was a year ago. It's also more fulfilling: 86% of developers say so, up from 76%.
9 hrs
spent per week learning AI or new technologies, up from 4 hours a year ago.
More than double the time investment in continuous learning in twelve months.
67%
report more time spent reviewing AI-generated code in the past year.
Review has become a larger share of developer time than writing new code from scratch.
60%
report more mental effort required to catch subtle bugs in AI-generated code.
The cognitive load of AI code review is rising, not falling, as AI adoption deepens.
58%
report increased accountability for code they didn't fully write.
Developers are the accountability layer for a growing volume of machine-generated work.
“In Q3 2025, 12% of developers said AI wrote half their code. Today it's 42%, and the decision to ship it still sits with one engineer, for far more code than it used to. That's why 78% of CTOs increased spending on review and validation. Every one of those dollars buys the same thing, which is an accountability layer that is currently one person deep.”
Darren Shimkus, CEO
What CTOs confirm, and where they don't
In a companion survey of 41 CTOs, the executive view of Q3 2026 largely matches what developers report from inside the work, with one notable exception.
78%
estimate a mid-level developer spends less than half the week writing new code.
Matches the developer side, where only 21% say they spend most of their week writing new code.
78%
increased spending on code review, QA, or validation infrastructure to support AI-generated code in the past year.
The developer oversight burden isn't just described. It's funded.
49%
increased hiring, upskilling, or budget for AI security in the past year.
As AI moves deeper into production code, executive investment is shifting toward the risks that come with it.
24%
invested in human skills over the past years less than in any technical AI speciality.
Yet human skills ranked as a top-three pay driver for developers whose earnings rose.
Research Assets
Executive Summary
A concise summary of the key findings and headline statistics on how AI oversight became the work. Built for sharing with leadership and press.
Data Deck · PDF
Complete survey data from developers and enterprise CTOs, including cross-tabs by experience level and region. The first edition to pair developer responses with CTO data. Ideal for reporters, analysts, and researchers.
Blog · Full Analysis
Full editorial analysis with findings on how developers spend their week, where CTO budgets are moving, and who answers for AI-generated code.
Previous Editions
A quarterly view into how software developers are adapting to AI, with each edition tracking the conversation as it evolves.
Q2 2026
Juniors and seniors see the AI shift in software engineering differently. That gap is changing how the next generation of developers learns the craft.
Q1 2026
1,329 developers across 61 countries on how AI is changing career trajectories and accountability in software teams. 64% of women say AI is expanding their career agency.
Q4 2025
501 developers and 19 project managers across 40 countries on how 2025 became the year AI shifted from threat to catalyst. 65% expect their role to be redefined.