On September 29, 2026, users around the world encountered problems accessing services provided by Anthropic, a company known for its Claude artificial intelligence tools. The affected services included Claude.ai, Claude Code, and Claude Cowork, which are used for tasks such as writing, coding, and collaboration. Anthropic confirmed the issue on its official status page at 4:21 p.m. in metropolitan France, acknowledging that users were experiencing "high error rates." These errors could lead to failed requests, loading problems, or prompts to reconnect. Some users found that retrying their actions later might resolve the issue. The outage was quickly noticed by users, with the tracking site Downdetector showing a sharp increase in reports shortly after 4 p.m. By 4:06 p.m., the site recorded 727 reports, and within a few minutes, the number had surpassed 1,500. However, the full scope of the disruption is still unclear, and no official estimate for recovery has been provided by Anthropic. Later in the day, Anthropic reported that it had implemented a mitigation strategy, which led to a "substantial" reduction in error rates. Despite this progress, some functions of Claude Code and Claude Cowork were still failing, and the company continues to work on a full resolution. The incident occurred just hours after the release of Claude Sonnet 5.5, a new model praised for its cost-effectiveness and performance. Anthropic is also reportedly preparing for its initial public offering, a process through which a private company offers its shares to the public for the first time. However, some reports have highlighted concerns about the company’s financial health, which may impact investor confidence during the IPO.