Claude Code’s source code appeared to have leaked online, exposing internal instructions and frameworks behind Anthropic’s AI agent.
The leaks appeared on GitHub, where copies and forks of the code were widely shared. Anthropic responded with copyright takedown requests, initially targeting more than 8,000 repositories before narrowing the scope to fewer than 100. The company says it wants to prevent competitors from cloning Claude’s features.
Claude Code is widely used by developers and businesses for coding and workflow automation. The leaks include details of Anthropic’s design approach, which gives competitors a clearer view of its strategy for building autonomous AI agents.
Anthropic fears its rivals could exploit the leaked code to copy Claude’s features. Industry analysts say the incident highlights the stakes in the race to develop agent‑based AI systems.
Also Read: OpenAI to show ads to ChatGPT Free and Go users
Protecting source code is now as important as advancing capabilities, especially as companies push toward more autonomous tools.
