Anthropic accidentally leaked the source code for its Claude Code AI agent this week, which went viral with millions of views and GitHub adaptations. Anthropic issued a DMCA takedown notice to control the spread, with a spokesperson stating, "We issued a DMCA takedown against one repository hosting leaked Claude Code source code and its forks." The incident carries irony, as Anthropic, OpenAI, and Google have faced lawsuits over using copyrighted material without permission. In September, a court ordered Anthropic to pay $1.5 billion in damages in a class-action lawsuit brought by authors and publishers. Reddit sued Anthropic last June for scraping user-generated content, and last month, Universal Music Group, Concord, and ABKCO filed a suit for illegally downloading over 20,000 copyrighted songs. Cybersecurity specialist Paul Price said the leak exposed the company's "harness" but noted, "It's more embarrassing than detrimental. Most of the real juicy stuff is in their internal source models and that wasn't leaked." An Anthropic spokesperson added, "We're rolling out measures to prevent this from happening again."