Sanction teardown · Administrative Review Tribunal, Australia · 2025-08-28
Butler and National Disability Insurance Agency
What happened
In Administrative Review Tribunal, Australia, a filing relied on an unnamed/unconfirmed AI tool to help draft legal argument. The court identified the following problems with the citations in that filing:
-
Fabricated (Exhibits & Submissions)Expert reports were prepared with AI assistance that introduced unverified or incorrect citations and inserted wording (e.g., the phrase 'minimum necessary'); Mr Reilly corrected some AI-generated citations and authors could not confirm citation accuracy.
Which AI tool
an unnamed/unconfirmed AI tool. Note: Charlotin's public database records tool attribution only where a court order, brief, or reporting on the matter states it explicitly; "unidentified" or "implied" means the record indicates AI use but does not name a specific product — we do not guess.
Outcome
Tribunal criticised the reliability of AI-assisted expert reports and gave them reduced weight
Additional detail
The Agency raised concerns that several expert reports were prepared with assistance of an artificial intelligence program and contained unverified citations and inserted text. The Tribunal noted admissions by at least one practitioner (Ms McPhee) that AI assisted drafting and found instances where citations had been corrected by another witness and where the author could not confirm whether AI added certain phrases. The Tribunal criticised the lack of independent verification and gave the reports reduced weight but did not impose sanctions.
How Citation Safe would have caught this
Citation Safe runs three deterministic layers before a brief is filed: (1) does the citation exist against CourtListener's database of published opinions, (2) if quoted, does that exact language appear in the source, (3) does the cited case actually support the proposition it is cited for. Fabricated case citations fail Layer 1. Fabricated or misattributed quotations fail Layer 2 even when the underlying case is real. Misrepresented holdings — a real case cited for a proposition it does not support — are the target of Layer 3. None of these checks involve asking another language model whether the citation looks right; they are lookups and text-matches against the actual source, which is why a hallucinated citation has to survive a direct lookup against the authoritative source — not another model's opinion — to earn a VERIFIED stamp; our measured false-verify rate is published live at /quality.
Check a brief before you file it → · See our live false-verify rate
Source: https://www.damiencharlotin.com/documents/760/Butler_and_National_Disability_Insurance_Agency_NDIS_2025_ARTA_1579_28_August_2025.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).