OPENAI'S MODELS 'MEMORIZED' COPYRIGHTED CONTENT, NEW STUDY SUGGESTS

Summary from the TLDR AI newsletter

A study suggested that OpenAI trained its models on copyrighted content without permission.