GitLab Is Cutting 30% of Its Countries and Betting the Company on AI Agents
GitLab's 'Act 2' restructuring means fewer countries, flatter management, and a bet that AI agents will write most code by 2028. The stock dropped 8.2%.
GitLab announced a restructuring on Monday that CEO Bill Staples called "Act 2." In plain terms: the company is cutting operations in up to 30% of countries, flattening its management structure, and betting that AI agents will write the majority of code within two years.
The market didn't like the uncertainty. GitLab stock dropped 8.2% after hours.
Staples laid out the reasoning in a memo that, for a corporate restructuring document, was unusually direct. "Software will be built by machines but guided by humans," he wrote. He said GitLab's current organizational structure was "built for a world where humans wrote the code." That world, in his view, is ending.
GitLab isn't the first company to make this pivot, but the scale is notable. This isn't a 5% workforce trim dressed up as an AI story. The company is physically exiting countries and collapsing management layers. For a company with roughly 2,200 employees, a 30% country footprint cut signals hundreds of roles affected across sales, support, and regional operations.
The strategy has a clear internal logic. If AI agents handle code generation, code review, and CI/CD pipeline management, then you need fewer people managing the workflow and more people who can build and supervise the agents themselves. GitLab is trading regional coverage for AI engineering density.
The risk is timing. AI coding tools have gotten good, fast. Claude Code and Cursor handle full features from a text prompt. Amazon's Q Developer injected code fixes directly into production pipelines at scale last quarter. The technology is real. But "AI agents write most code by 2028" is still a bet, not a certainty.
If the tools stall, or if enterprise adoption moves slower than Staples expects, GitLab will have thinned its operations for a future that didn't arrive on schedule. If the bet pays off, GitLab positions itself as the platform where AI-built code gets stored, reviewed, and deployed — and that's a business model with real leverage.
One detail worth noting: GitLab's restructuring follows a pattern. Microsoft, Google, and Amazon have all reorganized engineering teams this year to prioritize AI-native workflows. The difference is that GitLab is doing it publicly, with a timeline, and without the cushion of a trillion-dollar parent company.