[2/2] The State of DevOps Report 2026 Perforce (Category DevOps)

Continue. story on the new reportState of DevOps 2026From Perforce, I will share the main results that represent a managerial view of the relationship between DevOps practice maturity and AI. (These guys were picking up respondents so that they were mainly managers, making decisions on infrastructure and platform solutions in their companies.).

1️⃣ 70Percentage of respondents say DevOps maturity has significantly impacted AI success That's all. 38Percentage of organizations have actually built AI deep into several stages of SDLC 38% use it frequently but without standardization, and 17The percentage remains at the limited pilot level. The gap in maturity is huge: in high-maturity organizations 72Percentage of Leaders Say Deeply Embedded AI, in the Mid 43%, low - 18%. As a result, it is clear that for scaling AI initiatives, it is not the purchase of AI tools that is important, but rather the prepared ground in the form of mature engineering processes.

2AI Inherits Operational Model of Work, Not Fixes It In the report 32The percentage of organizations described as highly standardized 35% is mostly standardized and 34% continue to live in partial or chaotic delivery. That is, about a third of the market is still in the zone where the result depends on a particular team. Perforce calls this the problem of variation: as long as workflows, environments, and governance differ from team to team, AI will produce equally uneven results. Hence the emphasis on control plane: not “another tool”, but spherical templates, ballroom standards, spherical pipelines and manageable environments.

3️ There is already a confidence gap in the market between trust in AI and the actual integration of AI tools into processes. 77Percentage of respondents say they trust AI outputs 38The percentage actually built AI deeply into delivery, and 39% have fully automated audit data. In terms of measurement, the authors state the risk directly: organizations trust AI faster than they can build verifiability, auditability, and consistent measurability. This is an important antidote to the illusion that increasing local productivity in an IDE already means mature AI-native delivery.

4There is an economic effect, but it does not occur automatically. 74Percent of respondents believe that AI meets with high expectations. Separately, Perforce shows that ROI is stronger in those with a mature delivery system: high-maturity organizations on the market. 36% more often automated 61%+ Deployments from commit to sale and on 66% are more likely to "very effectively" respond to business incidents Low-maturity organizations, on the other hand, 78The percentage of delivery is not standardized. 19The percentage responds very effectively to incidents. If on the fingers, then without mature DevOps AI can speed up work, but at the same time increase the share of rework, variability of results, downtime time and costs.

In my opinion, the Perforce report confirms very well the basic thesis of my article.From PDLC to AI-native development“AI-native is not just another smart tool, but a redesign of the entire system for creating, verifying and delivering change.” Only I go further and reflect on the individual roles in this new process:)

#Engineering #AI #Metrics #Software #DevEx #Productivity #DevOps #Architecture #Culture #Engineering #ML #SystemDesign