Skip to content

Incident Tracker / United Kingdom

Harber v Commissioners for HMRC, [2023] UKFTT 1007 (TC)

·First-tier Tribunal (Tax Chamber)·UKDismissal

The first reported United Kingdom case of AI-fabricated authorities: before the First-tier Tax Tribunal, litigant in person Felicity Harber cited nine non-existent decisions generated by ChatGPT to argue a reasonable excuse for a capital-gains penalty. The tribunal found the cases fake, dismissed her appeal, and warned of the harm fabricated authorities cause.

Analysis
Harber v HMRC: the first reported UK case of AI-invented case law
Court
First-tier Tribunal (Tax Chamber)
Date
2023-12-04
Who used the AI
Not recorded
AI tool
ChatGPT
Conduct
A litigant in person submitted the names, dates and summaries of nine First-tier Tribunal decisions said to show a reasonable excuse for a late capital-gains disclosure; none of the authorities existed. The tribunal found they had been generated by AI.
Outcome
Dismissal
Penalty
No monetary penalty recorded
Regulatory outcome
No separate regulatory disposition recorded
Status
Verified against listed sources
Last checked
2026-09-11

What was fabricated or misused

  • David Perrin v HMRC (2019) — not a genuine FTT judgment
  • Jewell v HMRC (2016) — not a genuine FTT judgment
  • McMullen v HMRC (2018) — not a genuine FTT judgment
  • Milner v HMRC (2020) — not a genuine FTT judgment
  • Smith v HMRC (2021) — not a genuine FTT judgment
  • Oyesanya v HMRC (2020) — not a genuine FTT judgment
  • Baker v HMRC (2020) — not a genuine FTT judgment
  • Acheson v HMRC (2021) — not a genuine FTT judgment
  • Talal v HMRC (2019) — not a genuine FTT judgment

What the court said

“However, none of those authorities were genuine; they had instead been generated by artificial intelligence (“AI”).”
First-tier Tribunal (Tax Chamber), [2]
“Having considered all the points set out above, we find as a fact that the cases in the Response are not genuine FTT judgments but have been generated by an AI system such as ChatGPT.”
First-tier Tribunal (Tax Chamber), [21]
“We also find as a fact that Mrs Harber was not aware that the cases in the Response were fabricated, and did not know how to locate or check case law authorities by using the FTT website, BAILLI or other legal websites.”
First-tier Tribunal (Tax Chamber), [22]
“But that does not mean that citing invented judgments is harmless. It causes the Tribunal and HMRC to waste time and public money, and this reduces the resources available to progress the cases of other court users who are waiting for their appeals to be determined.”
First-tier Tribunal (Tax Chamber), [24]

Timeline

  1. Mrs Harber sent the Response with nine purported FTT authorities to the tribunal.
  2. At the first hearing, HMRC said it could not find some of the authorities.
  3. The tribunal reconvened and questioned how the cases had been found.
  4. The tribunal dismissed the appeal and found the nine cases AI-generated.

Why this case matters

Harber is the starting point for the UK tax-tribunal line of AI citation cases. It involved a litigant in person before domestic guidance had developed, and the tribunal had to rely on Mata v Avianca for an explanation of the harm caused by invented judgments. Later UK records in this tracker move the same problem into professional settings: Elden v HMRC involved regulated accountants, and Ayinde v Haringey involved lawyers facing Hamid jurisdiction scrutiny. Harber shows that good faith and lack of legal training did not make fabricated authorities harmless.

Practice note

A litigant or helper who uses a chatbot for legal research must obtain the full judgment from the FTT site, BAILII, The National Archives, or a recognised law-report source before citing it. A case name and a chatbot summary are not enough. The check should confirm the neutral citation, the issue decided, and whether the party actually won on the proposition being advanced.

Primary sources for this incident

Related regulation records for this incident

Read our full analysis of this case →

Similar incidents on the record

  • Folarin v Immigration Services Commissioner, [2026] UKFTT 135 (GRC)2026-01-29 · First-tier Tribunal (General Regulatory Chamber) · DismissalBefore the First-tier Tribunal (General Regulatory Chamber), immigration adviser Raphael Folarin, appealing the loss of his registration, cited multiple non-existent authorities obtained by asking ChatGPT for supporting…
  • MS (Professional conduct, AI generated documents) Bangladesh, [2025] UKUT 305 (IAC)2025-08-12 · Upper Tribunal (Immigration and Asylum Chamber) · Regulatory referralBefore the Upper Tribunal (Immigration and Asylum Chamber), barrister Muhammad Mujeebur Rahman cited a fabricated Court of Appeal case, Y (China), drafted using ChatGPT without verification. The tribunal, led by its Pre…
  • State ex rel. Oklahoma Bar Association v. Reeves, 2026 OK 372026-05-27 · Supreme Court of Oklahoma · WarningThe Oklahoma Supreme Court publicly reprimanded attorney Matthew Reeves under reciprocal-discipline rules after he admitted using ChatGPT to fabricate five citations in federal filings in Johnson v. Dunn — conduct alrea…
  • Rafique v Commissioners for HMRC, [2026] UKFTT 673 (TC)2026-05-06 · First-tier Tribunal (Tax Chamber) · DismissalBefore the First-tier Tax Tribunal, self-represented appellant Omar Rafique sent emails citing multiple non-existent authorities that Judge Bailey found were AI-generated and unreviewed. The tribunal refused to reinstat…

Cite this record

SafeLegalAI Global Legal AI Incident Tracker, “Harber v Commissioners for HMRC, [2023] UKFTT 1007 (TC)”, safelegalai.com/tracker/harber-v-hmrc (accessed 2026-09-16). Data: CC BY 4.0.

Report an issue with this record →