Leaked Claude code raises copyright questions for AI training and deployment (nytimes.com)

A leak of Anthropic's internal Claude testing code raises questions about copyright protections for AI model code and training materials in an era of rapid AI reproduction and deployment.