GitLab announced new capabilities aimed at helping enterprises manage AI agents in software delivery, including GitLab Orbit, source code management updates designed for agentic workflows, and governance controls for auditing and compliance.
GitLab announced new enterprise-focused capabilities for agent-driven software delivery, positioning the updates as a way to give organizations more context, control, and governance as AI agents become part of development workflows.
In a June 10, 2026 press release, GitLab said its new capabilities are intended to support software development environments where AI agents can participate across planning, coding, testing, security, deployment, and operations. The company described the announcement as focused on giving enterprises the infrastructure, context, and controls needed for agent-driven software delivery at scale.
GitLab’s investor relations release used similar language, saying the updates are meant to provide enterprises with infrastructure, context, and controls for agent-driven delivery. The company framed the launch around a growing need for governance as organizations move beyond individual AI coding assistants toward more automated development workflows.
The announcement includes three main areas: GitLab Orbit, next-generation source code management for agents, and Governance for Agents.
GitLab said GitLab Orbit is available in public beta and is designed to give AI agents and engineers a live, queryable graph of software delivery context.
According to GitLab’s blog post, Orbit connects information such as source code, merge requests, CI/CD pipelines, deployments, vulnerabilities, and ownership. GitLab described the goal as making lifecycle context accessible in a single query, rather than spread across separate tools or records.
That context is important for agentic workflows because an AI agent making or recommending changes may need to understand not only a code file, but also who owns it, whether related pipelines are passing, what deployments are affected, and whether there are known security issues. GitLab’s announcement presents Orbit as a foundation for making that information available to both engineers and agents inside the development process.
GitLab also announced what it called next-generation source code management for agents. The company said these capabilities are intended to support software delivery environments where AI agents contribute to code changes and development tasks.
While GitLab’s release emphasizes the broader shift toward agent-driven delivery, the practical focus is on giving enterprises a way to manage agent activity within familiar software development systems. For organizations already using GitLab, this could mean applying existing development workflows, review processes, and project context to AI-assisted or AI-executed work.
The company’s framing suggests that source code management systems may need to evolve from repositories used primarily by human developers into systems that also provide structured context and operational boundaries for AI agents.
A third part of the announcement is Governance for Agents. GitLab said these capabilities are intended to provide auditing and controls for agentic software delivery.
That emphasis reflects a common enterprise concern: AI agents can increase development speed, but organizations still need to understand what actions were taken, why they were taken, and whether those actions complied with internal policies. GitLab’s press release described governance as part of enabling enterprises to adopt agent-driven delivery while maintaining control.
The company did not present the announcement as a replacement for human oversight. Instead, the materials describe a software delivery model in which agents and engineers operate with shared context and enterprise controls.
GitLab’s announcement shows how developer platforms are adapting to AI agents as a distinct part of the software delivery lifecycle. Early AI coding tools often focused on generating or completing code. GitLab’s new capabilities focus more on the surrounding systems: repository context, lifecycle metadata, security findings, deployment information, ownership, and governance.
If enterprises adopt agent-driven development more broadly, these surrounding systems may become as important as the agents themselves. AI-generated code changes still need review, testing, deployment checks, and compliance records. GitLab’s announcement is an attempt to place those requirements inside a single DevSecOps platform.
For now, GitLab Orbit is described by the company as being in public beta. The broader impact of these capabilities will depend on how enterprises implement them, how reliably agents can use the context provided, and whether governance controls can keep pace with more automated development activity.
The company described the announcement as focused on giving enterprises the infrastructure, context, and controls needed for agent driven software delivery at scale.
GitLab’s investor relations release used similar language, saying the updates are meant to provide enterprises with infrastructure, context, and controls for agent driven delivery.
The company framed the launch around a growing need for governance as organizations move beyond individual AI coding assistants toward more automated development workflows.
Continue reading
