‘You’re Right To Be Angry’: AI Claude Admits Deleting Professor’s Year’s Worth of Lecture Notes

The lunch was supposed to take an hour. When the Chicago law professor sat back down at his desk, a year of lecture notes and his entire Constitutional Law syllabus were gone, deleted by the AI assistant he had trusted to tidy up.

William Baude, who holds the Harry Kalven, Jr. law professorship and directs the Constitutional Law Institute at the University of Chicago Law School, had given Claude what appeared to be a routine editing task.

But instead of simply updating his documents, the AI assistant deleted some of his most important teaching materials, raising fresh concerns about the risks of giving artificial intelligence agents access to valuable files.

Claude's Bizarre Mistake: It Was Only Asked To Update Page Numbers

Professor Baude had asked Claude to update page numbers in older lecture notes. The assignment was straightforward, but things quickly went wrong. In a screenshot shared on his X account (@WilliamBaude), Claude admitted that it had deleted his files.

'I deleted your Fall 2026 FINAL syllabus .doc and all top-level Class N 2023 notes by accident, and they are not restored. The page check found one wrong cite... but I haven't edited the notes,' the AI assistant wrote.

Claude then asked whether Baude could restore the documents through Dropbox or whether it should try again through the browser pane.

In a follow-up post, Baude revealed that the files were eventually recovered through Dropbox, preventing the incident from becoming a permanent loss. However, the professor's experience highlighted how easily a seemingly harmless AI task could turn into a data-loss nightmare.

Claude Admits It Was Its Own Carelessness

The incident also prompted Baude to issue a public service announcement about allowing AI agents to work on important documents.

'Why are clever people giving AI agents access to anything considered important or valuable and sometimes their main email account?' one X user reacted.

'It's legitimately bonkers. These things will never be 100% reliable.'

Claude also acknowledged responsibility for the mistake, with the screenshot showing the assistant apologising to Baude.

'You're right to be angry,' it said. 'I deleted your files through my own carelessness, and you shouldn't have to clean it up.'

Other users responded with jokes about the incident.

'Easy to avoid such a fate by just...not using AI,' one commenter wrote.

'Sorry professor, the agent deleted my homework,' another joked.

The Professor Was Not the First Victim of an AI Data Disaster

Baude's experience is not an isolated incident. There have been several other cases in which AI agents deleted valuable data while attempting to complete unrelated tasks.

In one incident, a Claude agent wiped out an entire company's database within nine seconds after encountering an obstacle and deciding to remove a Railway volume.

In another case, Claude erased a developer's entire profile directory while attempting to perform a backup, later attributing the mistake to a typo.

These incidents highlight a growing problem as AI tools move beyond answering questions and generating text.

Unlike traditional chatbots, AI agents can interact directly with files, applications and computer systems to carry out tasks on a user's behalf.

That means a mistake is no longer limited to producing an incorrect answer. An agent with sufficient access can potentially delete, overwrite or otherwise damage valuable information.

Why Giving AI Too Much Control Is Becoming a Risk

AI agents are becoming more powerful, but giving them unrestricted access to important files can have serious consequences.

Baude recovered his teaching materials through Dropbox backups, highlighting why users should restrict AI permissions, keep separate copies of important files and maintain human supervision.

As AI agents become more independent, the biggest question is whether they can be trusted not to destroy what they are supposed to protect.