The Seven Professional Sins of AI
The patterns. The consequences. The preventable ones.
These are the seven ways professionals get burned by AI. Not hypothetically. Actually. Each one has a name, a pattern, and a body count. Learn them before they find you.
Hallucination
The model fabricates with full confidence.
Confidentiality Failure
You pasted it in. The terms said what they said. The data left the building.
Over-Reliance
The AI gave you a confident answer. You accepted it. The judgment you were paid for was the part you skipped.
Context Collapse
The conversation grew. The founding facts drifted. The model kept answering. You never noticed it was working from a different foundation.
Prompt Naivety
Vague in, vague out — dressed up in fluent prose. The model answered the question you asked. That was not the question you needed answered.
The Competence Gap
You used a tool you do not understand on work you are responsible for. You cannot evaluate the output. You signed it anyway.
Auditability Failure
The examiner arrived. Show your work. The prompt. The model version. The date. What you did to verify it. You have the output. You have nothing else.
Everything else on this page assumes a system that begins each conversation empty. That assumption has expired. All four major vendors now carry a file between chats — assembled from your past conversations by a summarizer, held somewhere you cannot fully audit. It is not an eighth sin. It is a change in the conditions under which several of these seven occur, and it falsified two pages on this site, one of them a sin page below.
What that file contains, who decided what went in it, and why you cannot reliably remove things from it. →Nippon Life Insurance Company of America v. OpenAI Foundation and OpenAI Group PBC, N.D. Ill. No. 1:26-cv-02448, filed March 4, 2026.
A disability claimant uploaded her attorney’s letter to ChatGPT and asked if she was being gaslighted. ChatGPT said yes. It then drafted her motions, generated her legal arguments, conducted her legal research, and helped her file them in federal court. One filing cited a case that does not exist. By the time her insurer filed this lawsuit, she had submitted 74 documents across two docketed lawsuits. The insurer’s cost of defending a settled case: approximately $300,000.
Sin I: the fabricated citation. Sin II: privileged legal correspondence uploaded to a public model. Sin III: legal judgment outsourced entirely to an AI. Sin IV: a conversation that drifted so far from reality that the model was litigating a case that had already been dismissed. Sin V: prompts that produced the answers the user wanted to hear. Sin VI: a tool used for legal work by someone with no ability to evaluate its legal output. Sin VII: no record, no chain of custody, no way to reconstruct what the machine said or why.
The complaint states: “ChatGPT is not an attorney. Although it was able to pass the Uniform Bar Examination with a combined score of 297, it has not been admitted to practice law in the State of Illinois or in any other jurisdiction within the United States.”
This is what several of the seven sins look like in a single case.
Read about it →I would say I hope you find this useful. I have learned not to hope for things.
Reviewed