GitHub Reliability Declines, Anthropic Shifts Policies Amid Industry Shakeup
Review the recent GitHub reliability drop and Anthropic's policy changes to gauge impact on your AI workflows.
Review your CI/CD pipelines for GitHub dependency and consider alternative hosting if reliability is critical.
Summary
GitHub's reliability has slipped from a 90% uptime rating last month to a mere 86% this month, according to third‑party trackers, and a recent data‑integrity incident has pushed the platform into a zero‑nines zone. The company attributes the degradation to a 3.5‑fold increase in service load, though some observers suspect internal mismanagement. Meanwhile, prolific open‑source contributor Mitchell Hashimoto has announced his departure from GitHub, citing the platform's unsuitability for professional work.
Anthropic has taken a sharp turn, quietly nerfing Claude Code, banning corporate use of Claude, and raising prices, all of which erode trust in the model. These changes arrive just as the company is in what some call its “extraction” era, focusing on revenue over service quality. The broader industry is also reacting, with GitHub Copilot price hikes, Codex growth, and Google scrambling to build a competitive coding model.
The article highlights a trend toward higher costs and reliability concerns in AI‑centric tooling, underscoring the need for teams to reassess their dependencies on these platforms.
Key changes
- GitHub uptime fell from 90% to 86% and then to zero‑nines (86%) this month
- A data‑integrity incident caused further outages and reliability loss
- GitHub attributes the degradation to a 3.5‑fold increase in service load
- Mitchell Hashimoto, a prolific open‑source contributor, has quit GitHub over its unsuitability for professional work
- Anthropic has silently nerfed Claude Code, banned corporate use of Claude, and raised prices
- The industry is seeing price hikes for GitHub Copilot, explosive growth at Codex, and Google developing a new coding model