A judge in Argentina used generative AI to write part of a criminal sentence without proper supervision or transparency, leading to the annulment of the conviction by an appeals court.
On June 4, 2025, Judge Carlos Rogelio Richeri of the Criminal Court in Esquel, Argentina, issued a sentence condemning defendant Raul Amelio Payalef to two years and six months of effective prison for simple robbery. The Criminal Appeals Chamber of Esquel later discovered that Judge Richeri had used generative artificial intelligence (presumably ChatGPT) to draft part of the judicial decision without proper human supervision or transparency. The AI use was revealed when the judge accidentally included a phrase in the sentence that read 'Here you have point IV re-edited, without citations and ready to copy and paste.' The appeals court, composed of judges Hernan Dal Verme, Martin Eduardo Zacchino, and Carina Paola Estefania, determined that the use of AI violated due process by preventing control over the traceability of judicial reasoning and failing to guarantee human authorship of the decision. The court found that the judge had used AI specifically for resolving nullity issues regarding two essential pieces of evidence in the case. The appeals court annulled both the sentence and the preceding trial, ordering that the process be conducted again with a different magistrate, and referred the matter to the Superior Court of Justice of Chubut for ethical and disciplinary investigation.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Users anthropomorphizing, trusting, or relying on AI systems, leading to emotional or material dependence and inappropriate relationships with or expectations of AI systems. Trust can be exploited by malicious actors (e.g., to harvest personal information or enable manipulation), or result in harm from inappropriate use of AI in critical situations (e.g., medical emergency). Overreliance on AI systems can compromise autonomy and weaken social ties.
Human
Due to a decision or action made by humans
Unintentional
Due to an unexpected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed