EMA's survey on AI in DevOps covers the impact of AI on DevOps, vibe coding trends, and the need for better governance and security.
MINNEAPOLIS, October 23, 2025 — Perforce Software, the DevOps company for global teams seeking AI innovation at scale, announced new research showing that while AI is rapidly reshaping roles in DevOps, most enterprises struggle with governance. 62% of the IT leaders surveyed cited security and privacy risks as their top concerns, according to the report ‘AI in DevOps: Adoption Outpaces Governance and Changes the Role of the Developer” by Enterprise Management Associates (EMA) and sponsored by Perforce. The research is based on a survey of 206 IT executives and technical leaders.
AI Creates a Shift-Up Mentality for Developers
The study reveals a significant shift in developer roles, with professionals increasingly focused on oversight and validation activities rather than traditional coding tasks. According to the research, 57% of developers now spend more time on code review, standards, and quality oversight, while 53% report greater involvement in security, policy and compliance responsibilities. Additionally, 52% of developers dedicate increased time to validating AI-generated output.
“Human oversight has become a critical function in AI-augmented development environments, but beyond individual productivity gains, we also need to focus on applying AI to the interactions between different teams throughout the software delivery chain to achieve real value,” said Jake Hookom, EVP of Products at Perforce. “Scaling a shift-up mentality across multiple teams is how to connect AI investments to business goals.”
AI ROI Tracked in Engineering Teams with Emphasis on People and Process
Organizations primarily measure ROI through code quality and defect reduction (70%) and developer productivity (62%), with fewer tying AI benefits to release velocity or business outcomes—indicating ROI is still framed mostly in engineering terms rather than enterprise value. However, realized outcomes from the survey also indicated better people and processes through AI, including:
- 62% higher developer satisfaction
- 49% faster time to market
- 43% faster onboarding with junior developers
- 38% reduced toil in DevOps or pipeline tasks.
Duality of Impact Reveals Governance Needs for AI in Continuous Software Quality
Despite widespread AI adoption, the research identifies substantial governance gaps. While AI development tools deliver measurable benefits, including improved test coverage (56%) and more consistent code formatting (55%), a range of concerns confirms the need for governance:
- 52% report concerns about introducing security vulnerabilities, and 52% report concerns around code defects and bugs.
- 69% of respondents were concerned about overreliance on AI tools, while 61% expressed concerns about “blind faith” in AI results.
- Over half of the respondents (57%) report a negative or neutral experience with an AI tool hurting their workflow, primarily due to inconsistent results in code quality and testing.
Vibe Coding Seen as Beneficial, But Quality and Control Risks Persist
51% of those surveyed actively use vibe coding for new features, and most see strong flow benefits—but nearly half stress the need for oversight by experienced developers to correct inefficiencies and avoid structural risks. While adoption remains strong with vibe coding, only 3% of those surveyed felt like it would reshape developer workflows.
Vibe coding’s top benefits according to respondents
- Enhancement of productivity (cited by 38%)
- Helpful bridge to lower barriers for new or returning developers (cited by 33%)
- Useful for business users and less relevant with developers (cited by 20%).
Control and Quality are the Top Challenges around Vibe Coding:
- Risk of poor quality or insecure code (54%)
- Overreliance on inexperienced developers (48%)
- Lack of integration with existing DevOps/CI/CD pipelines (45%)
- Limited control over generated architecture or design choices (44%)
- Unclear IP ownership or generated applications (33%)
Enterprises’ Investment Focuses on Closing Interoperability Gap, Addressing Tool Sprawl, and Orchestration Challenges and Opportunities
“Beyond security risks, the biggest barriers to AI in DevOps are tool sprawl, interoperability gaps, and the difficulty of orchestrating AI assistants across workflows,” said Dan Twing, President and COO of EMA. “Without unified governance and vendor ecosystems that simplify integration, AI expansion will remain fragmented — limiting both scalability and return on investment.”
Key Factors that lead to these conclusions include:
- 48% “Strongly supportive” and 27% “Somewhat supportive” of using non-approved tools. Demonstrates decentralized experimentation and fragmented control leading to tool sprawl and lack of centralized control.
- 45% cite security concerns, 32% IP protection, and 25% vendor conflict when it comes to why organizations are blocking AI usage, which highlight integration and interoperability barriers during production rollout.
Survey respondents listed the below as the top improvements, they want with their AI-powered development workflows:
- 55% indicated a need for real-time vulnerability detection
- 53% wanted automated test generation
- 46% enhanced DevOps pipeline orchestration
- 46% wanted AI assisted performance testing or environment simulation
Perforce acted as a sponsor of the EMA report. For more details around the report’s findings, please download the report here and register for EMA’s upcoming webinar here.