How AI is Revolutionizing Software Development: Insights from DORA 2025

Google Cloud's DORA (DevOps Research and Assessments) may be less known in China than Gartner, but globally DORA is the most influential report in the DevOps field. Since 2014, DORA publishes an annual industry report based on surveys of thousands of practitioners, analyzing differences between high‑performance and low‑performance teams.

The 2025 DORA report surveyed nearly 5,000 technology professionals worldwide and found that AI is no longer a novelty but a near‑ubiquitous part of developers' toolkits.

Massive AI Adoption Drives Productivity Gains

The report shows that AI adoption among software professionals has surged to 90%, a 14% increase from the previous year, with developers and product managers spending about two hours per day using AI.

Dependence on AI varies by task type:

Writing new code: 71%

Modifying existing code: 66%

Writing documentation: 64%

Creating test cases: 62%

Explaining concepts: 62%

Analyzing data: 61%

Debugging: 59%

Understanding technical docs: 59%

Creating specifications: 56%

Code review: 56%

Maintaining legacy code: 55%

Most respondents (65%) heavily rely on AI in software development, 37% rely on it to some extent, and 20% rely on it heavily; 8% rely on it extremely heavily.

Key advantages reported:

AI adoption rate : 90% of software professionals now use AI in their work.

Significant productivity boost : Over 80% say AI has increased their productivity.

Reliance level : 65% depend on AI at least to some degree.

Code quality : 59% say AI positively impacts code quality.

Trust vs. Productivity Paradox

Despite widespread use, many developers remain cautious. The report finds a “trust paradox”: while 24% express high or strong trust in AI, 30% have low or no trust, indicating AI is viewed as an assistive tool rather than a full decision‑maker.

AI can improve individual performance, but its organizational impact is complex. The study shows AI use correlates with higher software delivery volume, a positive shift from previous years, yet ensuring software reliability before release remains a challenge.

Reflections on Team Workflows

The research suggests AI acts as a “mirror and amplifier”: it enhances efficiency in cohesive teams but can expose problems in loosely organized ones.

To better understand these dynamics, DORA identified seven team types—from “high‑performing harmonious” to “stuck in legacy bottlenecks”—providing nuanced insight into how AI success relates to performance, satisfaction, and work environment.

For organizations preparing to adopt AI, new tools can improve workflows, delivering both productivity gains and transformational benefits.

AI Application Blueprint: DORA AI Capability Model

DORA also introduced a new blueprint outlining seven core capabilities that amplify AI’s impact. The AI Capability Model, based on extensive research, pinpoints the combination of technical and cultural factors critical for success.

The model offers data‑driven guidance for AI adoption, emphasizing that realizing AI’s transformative potential requires changes in culture, processes, and systems to support the new era of software development.

References

[1] DORA AI Capability Model: https://cloud.google.com/blog/products/ai-machine-learning/introducing-doras-inaugural-ai-capabilities-model