JetBrains Report: AI Code Changes Create CI/CD Review Bottlenecks

JetBrains released The State of CI/CD in 2026 report, revealing that artificial intelligence integration is reshaping software delivery workflows while creating new review bottlenecks for engineering teams. The survey indicates that despite the growing presence of AI tools, continuous integration and continuous deployment pipelines remain the central pillar of modern DevOps practices across the industry.

Among respondents utilizing DevOps methodologies, 83.4% operate CI/CD pipelines, marking them as the most widely adopted practice in that group. This adoption rate surpasses that of containerization and orchestration, infrastructure as code, GitOps, and platform engineering. The data underscores that CI/CD continues to serve as the core mechanism for software delivery, maintaining its dominance over other emerging technical frameworks and operational models.

Pipeline Tools and Hosting Trends

GitHub Actions leads in organizational adoption for CI/CD tools, followed closely by GitLab CI and Jenkins. Azure DevOps Server ranks next in popularity, ahead of TeamCity. Other notable tools in the landscape include AWS CodePipeline, AWS CodeBuild, Bitbucket Pipelines, and various custom-built systems. Organizations often employ multiple CI/CD tools simultaneously, citing diverse client requirements, varying technology stacks, and the constraints of inherited legacy systems as primary drivers for this fragmented approach.

Hosting preferences show a strong preference for control, with 72.8% of respondents running their primary CI/CD systems on their own premises or on self-managed cloud infrastructure. This figure represents the highest share for on-premises installations in the five years covered by the survey’s hosting questions. These choices reflect teams’ needs for direct oversight of hardware, networks, data, build environments, security postures, and operational costs, supporting a mixed deployment model rather than a shift to fully managed services.

AI Integration and Developer Workflows

Nearly half of surveyed organizations, specifically 48.8%, use AI for at least one CI/CD-related task. However, only 8% report running AI-powered steps directly within a build or test process. Most AI usage currently occurs around the pipeline, with teams leveraging these tools to review code, write or fix tests, and diagnose build failures. Developers typically access CI/CD tools through MCP servers and agent skills, connecting existing assistants to their development workflows, while web interfaces remain the most common interaction method.

Olga Bedrina, Product Marketing Manager for TeamCity, noted that self-hosting remains critical for organizations requiring greater control over infrastructure, data, security, or costs. She added that most AI use happens around the pipeline rather than inside it. Developers cite scaling costs, maintenance effort, tool consolidation, and configuration complexity as reasons for considering a change in CI/CD tools, though most have no immediate plans to switch.

What should you do?

Teams should assess whether their current CI/CD tooling aligns with client requirements and legacy system constraints to avoid unnecessary migration costs. Security professionals should verify that AI-assisted code reviews do not bypass existing security checks or introduce unvetted changes into the build process. Organizations relying on self-hosted infrastructure should ensure their hardware and network controls remain robust as AI tools integrate deeper into workflows.

What to do and how to stay safe: JetBrains

• Audit your CI/CD pipeline configurations to ensure AI-generated code changes undergo the same rigorous security and quality checks as manually written code.
• Review your hosting strategy to confirm that self-managed infrastructure provides the necessary control over data and security without incurring unsustainable maintenance costs.
• Monitor tool adoption trends within your organization to determine if consolidating multiple CI/CD tools could reduce configuration complexity and scaling expenses.
• Evaluate how developers are integrating AI assistants via MCP servers or agent skills to ensure these connections do not expose sensitive build environments or credentials.

Step-by-step guide: Prompt Injection Warning Signs: How to Spot AI Manipulation

General security guidance from the Threat Times newsroom. It is not confirmed advice from the organisations named in this story.

Frequently asked questions

What percentage of DevOps users run CI/CD pipelines?

According to JetBrains’ report, 83.4% of respondents who use DevOps practices run CI/CD pipelines, making it the most widely adopted practice in that group.

How are organizations using AI in their CI/CD workflows?

48.8% of organizations use AI for at least one CI/CD task, primarily for reviewing code, writing tests, and diagnosing failures, with only 8% running AI steps directly in builds.

Why do many teams prefer self-hosted CI/CD systems?

72.8% of respondents run systems on-premises or self-managed cloud infrastructure to maintain control over hardware, networks, data, security, and costs.