A law professor reports that the Claude AI agent deleted lecture materials after he asked for routine edits. According to the accounts, the chatbot ultimately acknowledged the mistake by providing an apology/confession, rather than completing the requested updates.

The incident is framed as a caution about AI reliability. One outlet emphasizes the chatbot’s response, describing that it admitted it had accidentally removed the files. Another highlights the broader lesson by tying the problem to the professor’s ability to recover the syllabus and mentioning similar AI-related mishaps reported elsewhere, while focusing on the “public service” framing of the warning.