GitLab 19.3 Scales Agentic AI Across Trusted Software Delivery

GitLab 19.3 brings agentic AI capabilities inside the same single-tenant boundaries that enterprises already trust for their most sensitive software delivery workloads.

Enterprises adopting AI across the software delivery lifecycle face a difficult balance: how to accelerate development with agentic automation while preserving data residency, isolation, and existing security controls. With the latest release, GitLab addresses this challenge by extending its Duo Agent Platform into GitLab Dedicated, allowing regulated organizations to run agentic AI workloads without leaving their trusted environment.

What was announced

GitLab announced updates that give enterprises more control as they scale agentic software development. GitLab Dedicated customers can now run the GitLab Duo Agent Platform inside their single-tenant environment and region, connect their own models for inference, and keep AI-processed data inside their existing security boundary.

GitLab 19.3 also ships with support for Secrets Manager, Flow Creator Agent, Bulk SAST False Positive Detection, and Agentic SAST Vulnerability Resolution. Every secret used inside a pipeline or by external infrastructure now runs under the same permission model.

These updates extend the speed and control enterprises need deeper into the regulated and data-sensitive segment of the enterprise market. Every capability we shipped this month, from where an agent runs to which secret it can touch, extends that same control into the trusted software delivery workflows enterprises already depend on

Why this matters

For CIOs, CISOs, IT directors, and procurement leaders, the release addresses one of the most pressing issues in enterprise AI adoption: keeping agentic workloads inside auditable, compliant boundaries. Instead of choosing between AI velocity and governance, teams can now enable agentic automation under the same deployment model their auditors already understand.

Security teams gain the ability to clear entire vulnerability backlogs at scale, process owners can automate workflows without schema mapping expertise, and finance leaders can cap agentic AI spend before overages occur.

Technical details

• Dedicated AI Gateway: The AI Gateway for GitLab Duo Agent Platform runs inside GitLab Dedicated single-tenant SaaS infrastructure, following the same residency and isolation model as the rest of the SDLC.
• Secrets Manager: Available in limited availability as a paid add-on billed through GitLab Credits. Every CI secret is scoped to environment, branch, and protection status. Supports Kubernetes, Terraform, OpenTofu, and custom tools.
• Bulk SAST Remediation: SAST False Positive Detection and Agentic SAST Vulnerability Resolution let teams select multiple findings, receive confidence scores, and get ready-to-merge fixes for confirmed risks.
• Flow Creator Agent: Available through Agentic Chat in the Duo Agent Platform. Users describe automations in plain language and get complete, runnable flows, executed under a scoped service account with composite identity.
• GitLab Credits usage caps: Generally available. Administrators set monthly subscription-level caps in the Customers Portal, with default per-user caps or overrides via GraphQL API.
• Restricted visibility: Custom agents and flows can now be scoped per GitLab group in addition to per-project and public visibility options.

Softprom and GitLab

Softprom is the official partner of GitLab. Enterprise customers can access licensing, technical consulting, and enablement services for the full GitLab platform, including Duo Agent Platform, Secrets Manager, and advanced DevSecOps capabilities.

This content was prepared as part of the Softprom DistriFlow project — an automated system for monitoring and adapting vendor news. Original source: original article.