How GitLab scaled AI adoption with a federated governance model

GitLab developed a federated AI governance model combining centralized oversight with decentralized execution to accelerate adoption while maintaining quality. Their Enterprise AI team sets security standards and manages core vendor tools, while embedded AI Transformation Owners identify function-specific automation opportunities. A network of departmental AI Champions promotes best practices through hands-on guidance. To build fluency, GitLab's Talent Development team created the AI Literacy Ladder - a self-assessment tool generating personalized upskilling paths. Engineering-focused modules cover practical workflows like code review and pipeline fixes. Initial workshops saw 87% of technical attendees reporting immediately applicable skills, with a 22.3% increase in daily usage of internal AI coding tools within a month. The program tracks adoption through three metrics: training reach, progression through learning paths, and applied value measured by workshop feedback. Early results validate the hybrid approach - centralized guardrails prevent compliance risks while distributed experimentation drives organic adoption. Key lessons include co-locating technology and HR teams from day one, prioritizing judgment skills over tool-specific training, and treating enablement as a continuously updated product rather than a one-time initiative.