A popular artificial-intelligence app called Claude Code has been leaked after the company Anthropic accidentally exposed its underlying instructions. The leak was contained by taking down over 96 copies of the raw code from GitHub, a programming platform. A spokesman for Anthropic said there is no customer data at risk and that the leak was due to human error, not a security breach. To prevent similar incidents, Anthropic plans to implement new measures to secure its codes.
Source: https://www.wsj.com/tech/ai/anthropic-races-to-contain-leak-of-code-behind-claude-ai-agent-4bc5acc7