In the wild
When nobody checks the AI.
Consultancies, law firms, and newsrooms have all shipped AI-generated content with fabricated citations — in public, under their own name.
1,536 legal cases caught relying on AI-hallucinated content — and the tracker grows daily.
Deloitte Australia
Refunded a government report after a researcher found AI-fabricated citations and an invented quote from a Federal Court judgment.
Read the report →Deloitte Canada
A million-dollar report for a provincial government was found citing AI-generated research that does not exist.
Read the report →EY Canada
Withdrew a loyalty-fraud cybersecurity study after 16 of 27 citations were found hallucinated — including a McKinsey report that does not exist.
Read the report →Goldberg Segalla
A law firm was sanctioned after a lawyer filed ChatGPT-fabricated citations in a lead-paint case and never checked the work.
Read the report →California appeal
An attorney was fined $10,000 — 21 of the 23 case quotations in the opening brief were fabricated by AI.
Read the report →Mata v. Avianca
The case that started it all — New York lawyers sanctioned $5,000 after submitting six nonexistent court cases invented by ChatGPT.
Read the report →Chicago Sun-Times
A syndicated “summer reading list” ran ten books that don't exist — fake titles attributed to real, award-winning authors.
Read the report →Butler Snow
Attorneys defending Alabama's prison system were sanctioned and removed from a case for filing ChatGPT-fabricated citations.
Read the report →Don't let the next one be yours.
Argus audits AI-generated content for fabricated citations and false claims before you sign off.