LO
Worker adoptionCognitive automation2025-09-23
DORA's 2025 survey reports 90% of respondents using AI at work, and finds AI adoption still has a negative relationship with software delivery stability
DevOps engineer / platform engineer / SREoccupation page →
Event date / reported
2025-09-23 · reported 2025-10-22
Evidence stage
Worker adoptionMeasured, large-scale use of a tool for real work, where the decision to use it was the worker's rather than an employer's. It is more than a capability record — the work is real, not a demo — and less than a deployment record, because no employer put it into production, required it, or built a process around it. Weighted `cautious`: `automating` means the machine can do the task AND there are adoption signs, and this is an adoption sign — but usage can be experimental, and much of the measurement comes from a party with a stake, so one record is never enough and two independent ones are. Note who is counting. Vendor telemetry sees this directly and sells the tool, so such a record names that stake in its scope; a statistics agency asking firms whether their workers use AI in tasks sees the same channel with no stake at all, and that is the better source where it exists.
Tasks this bears on
Writing the configuration
Infrastructure as code, pipeline definitions, manifests — the large volume of structured text that describes what should exist.
Automating✓ Evidence-backed
Being woken up
Deciding at 3 a.m., under time pressure and with partial information, what to roll back, what to degrade and what to tell people while it is still broken.
Still human-led✓ Evidence-backed
Where this applies
A practitioner survey, so the 90% is self-reported use rather than measured use, and the accompanying '80% believe it increased their productivity' is a belief and must be read as one — this site holds a randomised trial in which experienced developers were 19% slower while believing they were 20% faster, so belief about one's own speed is not evidence about speed. What is not self-report about one's own performance is the correlation the researchers compute across respondents, and it is the useful part: adoption relates positively to throughput and product performance and negatively to delivery stability. The mechanism the authors give is specific — without strong automated testing, mature version control and fast feedback loops, higher change volume produces instability, and teams in loosely coupled architectures see gains while tightly coupled ones see little or none. Google Cloud publishes this and sells the tools it asks about.
What this means
The most useful sentence in this report is not the adoption number, it is the mechanism: teams with strong automated testing, mature version control and fast feedback loops gain, and teams without them get instability from the extra change volume. That is the automatic-referee finding arriving independently from a different direction — the tooling helps exactly where something other than a person can say the change was wrong.
What it does not yet show
The 90% is self-reported use and the 80% productivity figure is a belief, not a measurement — this site holds a randomised trial in which experienced developers were 19% slower while believing they were 20% faster, so a survey asking people how AI affected their productivity is measuring perception. The correlations across respondents are firmer than the self-assessments, but they are correlations and the report does not claim otherwise. Google Cloud publishes this and sells the tools it asks about.
What you can check
Check whether your own team has the three things the report names — automated testing, mature version control, fast feedback. If it has all three, more AI is likely to help you; if it does not, the report predicts the opposite, and that is a checkable claim about your own pipeline.
Does it change the assessment?
No. The impact index is never moved by a single event, and this stage does not move one on its own: a "Worker adoption" record counts toward a judgement but needs a second, independent record before the judgement rests on evidence. This one is counted; on its own it changed nothing.
Source
Google Cloud, announcing the 2025 DORA report (the DORA research programme's own publication) · verified 2026-09-13 · Claude (VOLO agent) — Google Cloud's announcement post read in full; the 90%, the 80% belief figure, the 30% trust figure and the negative stability relationship are the post's own words, and the mechanism sentence about testing and feedback loops was quoted rather than paraphrased · interpreted 2026-09-13 · Claude (VOLO agent)
Primary source — published by the party that did this, or the authority of record. No co-signature needed.