Sanction teardown · Federal Court (Canada), Canada · 2025-09-03
Karen Tyler v Attorney General of Canada
What happened
In Federal Court (Canada), Canada, a filing relied on ChatGPT to help draft legal argument. The court identified the following problems with the citations in that filing:
-
Fabricated (Case Law)Applicant relied on a non-existent authority cited for allowing extensions for CPPD/mental illness; Court found it does not exist and disregarded it.
-
Fabricated (Case Law)Applicant cited a purported decision excusing delay due to psychological illness; Court found the citation does not exist and disregarded it.
-
Fabricated (Case Law)Applicant relied on an authority allegedly granting extension for disability and self-representation; Court found the citation fabricated and disregarded it.
-
Fabricated (Case Law)Applicant cited a decision allegedly cautioning that procedural rules must not bar access to justice for disabled persons; Court found the citation non-existent and disregarded it.
-
Fabricated (Case Law)Additional 'hallucinated' citation in the Reply identified by the Court as not reliable or non-existent and therefore disregarded.
-
Fabricated (Case Law)Additional 'hallucinated' citation in the Reply identified by the Court as not reliable or non-existent and therefore disregarded.
Which AI tool
ChatGPT. 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
The Court disregarded the erroneous (fabricated) citations and noted the undeclared use of AI increased respondents' work; costs of 1,260.00 were awarded to the Respondent in part for this reason. (monetary penalty: 1 .)
Additional detail
The self-represented Applicant declared use of ChatGPT in her Reply and included multiple non-existent case authorities and other 'hallucinated' citations in her written materials. The Respondent identified several fabricated cases; the Court conducted its own checks, found the cited authorities did not exist, and held the Applicant breached her duty to verify AI-generated authorities. The Court disregarded the erroneous citations and awarded costs of $1,260 to the Respondent, noting the undeclared AI use imposed additional burden on the Respondent and the Court.
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/2767/Tyler-v.-Canada-Attorney-General-2025-FC-1457.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).