GitHub is experiencing significant reliability issues due to a 3.5x increase in service load, which has led to growing dissatisfaction among developers, including prolific open source contributor Mitchell Hashimoto who is quitting the platform. GitHub’s leadership attributes the degradation to this surge in demand, though some speculate the problems may be self-inflicted, according to pragmaticengineer.com.
The surge in load appears linked to the rising use of AI-powered coding tools like GitHub Copilot, which has seen dramatic price increases amid explosive growth in AI coding models such as Codex. Meanwhile, competitors like Anthropic have recently restricted access to their Claude Code AI model and raised prices, causing frustration in the developer community. Amazon has lifted its ban on Claude Code and Codex to improve its own AI coding agent, Kiro, highlighting the competitive dynamics in AI-assisted software development.
This situation matters because it underscores the challenges of scaling AI-driven developer tools while maintaining platform stability. GitHub’s reliability dropping below 90% contrasts with other vendors who have managed to handle increased AI workloads more effectively. The tension between rapid AI adoption and infrastructure capacity is reshaping how companies approach developer tools, pricing, and service quality in a market where small, focused AI teams are becoming the norm.
Looking ahead, the industry will be watching how GitHub addresses these reliability issues and whether it can stabilize service amid growing AI demand. The evolving strategies of major players like Anthropic, Amazon, and Meta in deploying AI coding assistants will also be key indicators of how the developer tools landscape adapts to these pressures, with potential impacts on pricing, access, and innovation.