Anthropic Leaks Claude AI Source Code
Anthropic PBC inadvertently released source code for its popular Claude AI agent, prompting inquiries into its operational security and leading developers to seek insights into the startup’s future intentions. “Earlier today, a Claude Code release included some internal source code. No sensitive customer data or credentials were involved or exposed,” an Anthropic spokesperson stated. “This was a release packaging issue caused by human error, not a security breach. We’re rolling out measures to prevent this from happening again.”
The leak of fundamental source code — the second incident in merely a week — sparked a conversation within the community regarding fresh insights into the functioning of Anthropic’s well-known coding agent. Developers stated on X that they were meticulously examining the details to understand how the startup planned to advance the platform. Numerous experts have expressed apprehensions regarding possible security vulnerabilities stemming from the unintended exposure.
The leak follows a report from Fortune revealing that the company inadvertently exposed thousands of files to the public, which included a draft blog post outlining a significant upcoming model referred to internally as both “Mythos” and “Capybara,” highlighting potential cybersecurity risks.
