Recent outages underscore Anthropic's infrastructure strain as Claude scales amid surging demand for its large language models. A major incident on September 29 disrupted claude.ai, the API, Code, and Cowork for roughly two hours with elevated errors, following a September 22 model-specific spike and an October 1 credit-balance delay that degraded API and console performance. These align with Anthropic's pattern of frequent short incidents in 2026, driven by high usage, model complexity, and billing system loads. Traders see the tight 5-7 versus 8-10 day spread as reflecting this moderate but persistent volatility, tempered by quick resolutions and competition from more stable platforms like OpenAI, while acknowledging that new releases or usage surges could shift outcomes.