GitHub experienced dual service outages in August 2026 caused not by code or configuration errors but by sheer capacity failure: monthly commits nearly doubled from 1.4 billion in April to 2.9 billion in August. The platform is now processing roughly 275 million commits per week, on pace for ~14 billion in 2026 versus 1 billion in 2025. GitHub Actions compute jumped from 500 million minutes per week in 2023 to 2.1 billion in a single week in 2026. Pull requests opened by AI agents surged from 4 million in September 2025 to 17 million by March 2026.
GitHub's initial capacity plan, launched in October 2025, targeted 10× growth. Within four months, the actual requirement was closer to 30×. The August 17 outage triggered a client-side retry loop in Copilot services that amplified recovery difficulty. GitHub's engineering team acknowledged that critical infrastructure components failed to scale with demand. The company is now pivoting to aggressive scaling, service isolation, and a migration to Microsoft Azure—all while simultaneously sustaining a 14× year-over-year load increase from agentic coding (Claude Code, Cursor, Copilot, Devin).
For any organization shipping AI coding agents, this is a live example of the gap between human and machine load profiles. GitHub handled human developer traffic for two decades; agentic agents create sustained, API-driven hammering at scales that break traditional capacity planning. The bottleneck in software is no longer writing code—it is reviewing, testing, securing, and governing agent output at volumes human teams cannot inspect in real time. For architects, the lesson: build guardrails, automated testing, and governance as aggressively as you scale output.