AI-driven coding and increased reliance on dependencies introduce significant application risk. Transitive dependencies are a major source of vulnerabilities, impacting a large percentage of software releases. GitLab's new Dependency Scanning Auto-Remediation feature aims to address this by automatically fixing vulnerable packages. When a vulnerability is detected, GitLab opens a merge request to update the dependency. If the update causes build-breaking changes, AI is employed to fix these issues, ensuring the pipeline passes. This process reduces security backlogs without diverting developer resources. High-severity vulnerabilities are fixed promptly, meeting compliance deadlines. Breaking upgrades are presented as merge requests ready for approval. The system upgrades vulnerable dependencies as they are found, preventing backlogs and maintaining compliance. Breaking changes are resolved by AI, allowing developers to review functional updates. All changes have an audit trail and require reviewer approval before merging. Automated version bumping addresses vulnerabilities, while AI handles breaking changes. This feature supports various ecosystems and is configurable to match risk tolerance. Remediation runs through existing pipelines, inheriting access controls and approval gates. The feature is currently in public beta and available on GitLab.com. Automated version bumping is included with GitLab Ultimate, while agentic breaking-change resolution is available through a free trial or with GitLab Credits.