JetBrains Report Reveals CI/CD Universal, Self-Hosted Resurgence, and AI's Peripheral Role in Pipelines

JetBrains has released its "State of CI/CD in 2026" report, drawing insights from multiple surveys including the 2026 CI/CD Tools Adoption Survey and the Developer Ecosystem Survey. The key findings reveal that CI/CD has become a ubiquitous practice, with 83.4% of developers actively running pipelines. This widespread adoption positions CI/CD as a fundamental component of modern software development, rather than a competitive advantage. The report also notes a significant trend towards self-hosted CI/CD solutions, with 72.8% of primary CI/CD systems being run on-premises or on self-managed cloud infrastructure, and on-premises installations reaching a five-year high at 30.5%. Another notable observation is the role of Artificial Intelligence: while nearly half of organizations (48.8%) leverage AI for some CI/CD tasks, only a small fraction (8%) integrate AI-powered steps directly into the build or test process. This report matters significantly to practitioners because it redefines the baseline for software delivery. With CI/CD being almost universal, the focus shifts from *whether* to adopt it to *how effectively* it's implemented and optimized. The resurgence of self-hosting indicates that many organizations are prioritizing control, data sovereignty, and potentially cost savings over the perceived simplicity of fully managed cloud CI/CD. This trend could be driven by increasing regulatory demands, security concerns, or a desire for greater customization and performance tuning. The limited in-pipeline AI adoption, despite broader AI use in development, suggests that the industry is still exploring how to best integrate AI into the core, performance-critical path of CI/CD. This trend aligns with the broader movement towards platform engineering and internal developer platforms, where organizations seek to provide developers with curated, self-service environments that offer both flexibility and governance. The desire for more control over the CI/CD stack, as evidenced by the rise in self-hosting, directly supports the platform engineering ethos of building robust, opinionated internal tools. Similarly, the cautious integration of AI into the build process reflects a common pattern in the adoption of new technologies in critical infrastructure: initial experimentation and use in less sensitive areas (like code review or test generation) before deeper integration into core workflows. This mirrors the gradual adoption of practices like GitOps and advanced observability, which started as niche approaches and are now becoming mainstream for enhancing reliability and efficiency. In practice, this means that practitioners should evaluate their current CI/CD strategy with a critical eye. If not already in place, establishing a robust CI/CD pipeline is no longer optional. For those already using CI/CD, the focus should be on optimizing existing pipelines, exploring toolchain consolidation to reduce complexity, and carefully considering the trade-offs between managed cloud services and self-hosted solutions. The data suggests that a self-hosted approach might offer greater long-term control and cost efficiency for some, especially larger enterprises with specific compliance needs. Furthermore, while AI's direct impact on build times might be limited today, practitioners should keep an eye on advancements in AI-driven testing, anomaly detection within pipelines, and intelligent resource allocation, as these areas are ripe for future innovation and could significantly enhance CI/CD efficiency and reliability. Experimenting with AI in peripheral CI/CD tasks can provide valuable experience and prepare teams for deeper integrations as the technology matures.