Northeastern University: Anthropic’s ‘anonymous’ interviews cracked by professor with an LLM. “A Northeastern University professor used a large language model to de-anonymize a subset of interviews conducted by Anthropic’s Interviewer tool.”