Judge investigated for misuse of AI to dramatically increase output
A Brazilian judge is under investigation for reportedly using unapproved generative AI tools to generate judicial decisions, resulting in a 12-fold surge in case output and sentences containing fabricated legal precedents. Judicial supervisors had previously warned him against using such tools, and a formal inquiry is underway.
What happened
Judge Tonny Carvalho Araújo Luz of the 2nd Court of Balsas County, Maranhão, Brazil, used generative AI tools not approved by judicial authorities to conduct legal research and generate judicial decisions.
His monthly output of sentences increased from an average of 80 per month in the first seven months of 2024 to 969 in August 2024—a 12-fold increase. Judicial inspectors noted that decisions from this period appeared to follow a single pattern, suggesting improper use of AI.
Reported impact
The judge's sentences reportedly included non-existent legal precedents, and cases were dismissed without resolution on their merits. Sentences were later reformed when challenged through appeals procedures.
Index analysis
In October 2024, two judicial supervisors from the Corregedoria recommended that Judge Luz avoid using open, non-approved generative AI tools for legal research, directing him to ensure decisions complied with applicable law and valid judicial precedents.
Governance controls
The Corregedoria-Geral de Justiça do Maranhão opened a formal inquiry to investigate the judge's conduct. The corregedor-general directed a deeper investigation to examine dismissed cases without merit resolution, reformed sentences, and those using fabricated precedents.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.