Incident Tracker / European Union
Raad van Discipline 's-Hertogenbosch, ECLI:NL:TADRSHE:2026:93 (Case 26-379/DB/LI/D)
The Raad van Discipline in 's-Hertogenbosch, a Dutch bar disciplinary council, formally reprimanded an advocate who used AI to generate ECLI case citations in two tenancy defences, several of which did not exist or referred to unrelated rulings. Rejecting a mere warning, it found breaches of the core values of expertise and integrity and ordered EUR 1,250 in costs.
- Court
- Raad van Discipline 's-Hertogenbosch (Netherlands bar disciplinary council)
- Date
- 2026-07-27
- Who used the AI
- Lawyer
- AI tool
- Not identified in the record
- Conduct
- On a complaint by the Dean, an employment-law advocate handling two tenancy cases was found to have filed defences containing AI-generated ECLI citations, some nonexistent and others pointing to unrelated divorce and criminal judgments. She later filed rectifications withdrawing them and also failed to attend hearings.
- Outcome
- Other
- Penalty
- €1,250
- Regulatory outcome
- Raad van Discipline 's-Hertogenbosch: Formal reprimand (berisping) on a complaint by the local Dean for breaches of the core values of expertise and integrity, with EUR 1,250 costs (EUR 750 to the Bar, EUR 500 to the State). (2026-07-27)
- Status
- Verified against listed sources
- Last checked
- 2026-09-03
What was fabricated or misused
- Dossier M defence cited eight decisions; one ECLI did not exist and seven ECLIs belonged to different decisions, including a divorce order and a criminal case
- Dossier A defence included incorrect references to partly nonexistent case law; the council did not enumerate them
- Dossier M paragraph 1 misstated the Landelijk Procesreglement by saying kanton pleadings had a fifteen-page maximum
- Dossier M paragraph 22 misused article 7:274(1)(c) BW as a main-residence termination rule; the council found that provision inapplicable to the primary defence
What the court said
“Voor een deel van genoemde verwijzingen gold dat het ECLI-nummer helemaal niet bestond. Voor een ander deel gold dat het ECLI-nummer verwees naar andere uitspraken dan die werden omschreven in de conclusie van antwoord.”
“Door op deze wijze gebruik te maken van een hulpmiddel, in dit geval een AI-tool, heeft verweerster de op haar rustende verantwoordelijkheid miskend.”
“Verweerster heeft dit vertrouwen met haar wijze van procederen geschaad en heeft daarmee niet alleen de kernwaarde deskundigheid, maar ook de kernwaarde integriteit en gedragsregel 8 (verstrekken van informatie waarvan de advocaat weet of behoort te weten dat deze onjuist is) geschonden.”
“Alles afwegend is de raad van oordeel dat een berisping passend en geboden is.”
Timeline
- Advocate filed the Dossier M defence with AI-generated ECLI references
- Rectification in Dossier M withdrew untraceable ECLI references
- Dossier M oral hearing proceeded without the advocate present
- Kantonrechter asked for an explanation of AI use and withdrawn citations
- Dean and advocate discussed AI use in Dossiers M and A
- Dean filed the disciplinary objection with the council
- Council imposed a reprimand and EUR 1,250 costs
Why this case matters
The Dutch decision is the tracker’s first EU-member bar disposition, and it differs from Mata v Avianca and LeDoux because the discipline came through a dean complaint rather than a court sanction motion. The council treated AI citations as part of ordinary quality-of-service and core-value analysis: the advocate lacked tenancy expertise, copied AI-generated cases without reading them, failed to attend hearings, and left clients to face judges after withdrawing the references. Beside Dayal, it shows a regulator can respond to hallucinated authorities with a public reprimand and costs even where confidentiality concerns are not proved.
Practice note
Before a Dutch advocate files AI-assisted research, each ECLI should be opened in an official database, read, and matched to the proposition in the pleading. If a correction withdraws citations, counsel should replace the legal support or explain the gap, then attend the hearing so the judge and client are not left with unsupported defences.
Primary sources for this incident
Related regulation records for this incident
No related regulation record has been linked to this incident yet.
Similar incidents on the record
- In re Brian E. Mitchell, USPTO Proceeding No. D2026-162026-07-27 · US Patent and Trademark Office, Office of Enrollment and Discipline · OtherThe US Patent and Trademark Office's Office of Enrollment and Discipline publicly reprimanded California patent attorney Brian Mitchell, whose AI-drafted claim-construction chart in a federal patent case cited nonexiste…
- Cervantes v. Bianco, No. S295866 (Cal. Sept. 24, 2026)2026-09-24 · Supreme Court of California · Regulatory referralIn California, the state Supreme Court disapproved the improper use of AI tools in the Riverside County sheriff's briefing in a ballot-seizure writ case, after the law firm misquoted Common Cause v. Board of Supervisors…
- Douglas v. Deutsche Bank National Trust Co., No. 24-CV-1099 (D.C. Ct. App.)2026-09-03 · District of Columbia Court of Appeals · Regulatory referralIn the District of Columbia Court of Appeals, Deutsche Bank's appellee brief cited four nonexistent authorities after counsel used Google's generative AI search tool without verification. The court struck the brief, ref…
- SRA v Abhishek Kumar, Case No. 12884-2026 (SDT)2026-08-25 · Solicitors Disciplinary Tribunal · Strike-off / disbarmentIn the United Kingdom, the Solicitors Disciplinary Tribunal struck Registered Foreign Lawyer Abhishek Kumar off the register after he used generative AI to file false citations and misleading quotations in his defence t…
Cite this record
SafeLegalAI Global Legal AI Incident Tracker, “Raad van Discipline 's-Hertogenbosch, ECLI:NL:TADRSHE:2026:93 (Case 26-379/DB/LI/D)”, safelegalai.com/tracker/raad-van-discipline-2026-93 (accessed 2026-10-07). Data: CC BY 4.0.