Marie Pia Fazio v Canada Revenue Agency
Federal Court · May 21, 2026
- Jurisdiction
- international
- Court
- Federal Court
- Record ID
- marie-pia-fazio-v-canada-revenue-agency-2026-05-21
- Observed outcome
- The Court dismissed the contempt allegation as unfounded and ordered the applicant to pay costs to the respondent.
- Known monetary consequence
- Not recorded
Why this matter warrants attention
The Federal Court of Canada issued an adverse costs order against a pro se litigant who made unfounded contempt allegations against opposing counsel. The Court rejected the litigant's attempt to blame artificial intelligence for the error, affirming that parties remain strictly responsible for the accuracy of their AI-assisted submissions under the Court's AI Notice.
The Court reasoned that the applicant was personally aware of the nature of her allegations, as evidenced by her specific citations to Rule 466(d) in prior correspondence. Furthermore, the Court held that the Federal Court's AI Notice requires a 'human in the loop' and mandates that the signing party bears full responsibility for the accuracy of all filings, regardless of whether AI tools were used to suggest the content.
This case establishes that litigants cannot use AI as a shield to avoid accountability for unfounded legal allegations. It reinforces the principle that the duty of candor and the responsibility for the accuracy of court filings rest solely with the human party, regardless of the technology used in drafting.
Failure modes and consequences
- Pro Se
- Fake Citations
- Administrative
- Trial
- Monetary
- AI attribution
- Admitted
- Recorded tool
- Unidentified
- Known monetary consequence
- Not recorded
- Procedural posture
- Judicial review of a Canada Revenue Agency decision, including a motion for costs and other relief.
Damien Charlotin case archive
Publisher document archive. The recorded document is hosted in the upstream publisher archive.
Open underlying sourceA recorded source link is not a substitute for checking the underlying order, filing, opinion, or disciplinary record.
What this record does—and does not—establish
The applicant admitted to using artificial intelligence in her submissions.
The record summarizes the outcome described in the linked public source. The recorded document is hosted in the upstream publisher archive. This page is not a substitute for the complete docket, subsequent history, or jurisdiction-specific advice.
Latest evidence review: Primary Document Verified (high confidence).
Controls suggested by the public record
- Implement mandatory human review protocols for all AI-generated legal arguments and citations.
- Ensure that all legal allegations, particularly those involving contempt or professional misconduct, are independently verified against the applicable rules of procedure before filing.
- Document the human oversight process for all AI-assisted submissions to demonstrate compliance with court-specific AI practice guidelines.