How are developers using AI? Inside our 2025 DORA report (AI column)
This article from 23 In September, I was given a Pulse by ChatGPT, which makes a selection of my interests. And I did have a lot of conversations with ChatGPT about DORA research and the impact of AI on software development - we at T-Bank even launched it. A large study of AI in the engineering culture of Russia. But if you go back to the DORA report for 2025 It is now available and an executive summary is available. article Google blog. In short, here’s how AI affects software development based on survey authors’ measurements.
1. The use of AI has become almost universal
- Around. 90% of developers use AI tools More. 80% believe that it increases their productivity
- Around. 59% believe it improves the quality of the code
- Large majority (65%) Respondents rely heavily on AI in software development: 37% report a “moderate degree” of dependence 20% - about "large" and 8% is about "significant."
2. The paradox of trust and productivity Despite this penetration, trust in AI is not so great.
- 24Percentage of respondents report ‘very high’ (4%) or "big" (20%) trust
- 30Percent trust him "a little bit" (23%) It turns out that respondents may find AI’s work useful even if they don’t fully trust it. They may have built processes so that AI works more like an assistant, but the work is not delegated to it.
3. Team archetypes This year’s study also found that AI can act as a “mirror and multiplier.” In cohesive organizations, AI increases efficiency. In fragmented - reveals weaknesses. To dig deeper into this, the authors uncovered seven different team archetypes ranging from “harmonious and successful” to teams facing “legacy bottlenecks.”
4. A blueprint for guiding AI in organizations In the survey report, the authors propose a “DORA AI Capabilities Model,” which identifies seven basic practices that pave the way for the effective application of AI.
#AI #Software #Engineering #Management