GitLab has announced GitLab 19.2 and included agentic AI abilities to assist developers in managing the increasing complexity of AI-generated code and dependencies as well as security challenges while still keeping the enterprise-level governance intact. The new version comes with Dependency Scanning Auto-Remediation in public beta where an AI agent is able to detect any vulnerable dependencies, create a merge request for the solution to the problem and fix any issues related to the build without interrupting the approval process and audit logs. In addition, GitLab has introduced Security Review Flow in public beta, where AI is used to scan the application’s logic and find any vulnerabilities like authorization, business logic issues, information disclosure and mass assignment problems.
Also Read: Cadence Launches AuraStack AI Super Agent to Transform PCB and Advanced Packaging Design
The feature provides severity ratings and suggested fixes while ensuring final approval remains with human reviewers. In addition, GitLab Duo CLI is now generally available, extending GitLab Duo Agent Platform capabilities to developers’ terminals with full project and pipeline context, allowing users to investigate unfamiliar code, troubleshoot failed pipelines, and propose fixes directly from the command line. In addition, the release takes advantage of research done by Forrester Consulting, where they found out that organizations deploying the GitLab Duo Agent Platform stand to enjoy a 400% ROI within six months. GitLab 19.2 takes advantage of AI automation combined with governance capabilities to facilitate fast delivery of applications that are secure in the DevSecOps lifecycle.