AI Hallucinations Hit the Courtroom: Lawyer Fined for Fabricated Precedents
Syracuse, Italy – A lawyer in Syracuse, Italy, has been sanctioned with a 30,000 euro fine after submitting four non-existent legal precedents to a court, generated by an artificial intelligence (AI) system. The incident underscores the growing risks associated with the uncritical adoption of AI tools in the legal profession.
The case, detailed in decision no. 338/2026 filed on February 20th, revealed that the judge found no record of the cited rulings within the Court of Cassation’s database. The lawyer had utilized a generative AI tool without verifying the accuracy of the information provided, leading to the introduction of fabricated legal references.
The court’s ruling imposed a triple penalty: over 14,000 euros in legal costs for the opposing party, an equivalent amount for aggravated liability related to the reckless dispute, and an additional 2,000 euros to the Fines Fund. The judge emphasized that AI language models are not reliable archives of jurisprudence, but rather probabilistic systems capable of generating plausible text without guaranteeing truthfulness.
While the referenced rulings did exist, they did not contain the passages cited by the lawyer and pertained to unrelated subjects. A simple verification process would have immediately identified the errors. The court deemed the lawyer’s omission a “serious fault,” establishing a precedent that will likely influence the future use of AI within the legal field.
Giuseppe Gurrieri, vice-president of the Criminal Chamber of Syracuse, noted that AI can be a valuable tool for legal analysis and document construction, provided it is used with “method, and control.” This case serves as a stark warning about the potential for AI to generate misleading content and the critical demand for professionals to verify AI-generated outputs against primary sources.
Worth a look