Gabinete RP #42: Advogado sancionado por usar IA?!

The Court of Appeal of Guimarães sanctioned a lawyer for using AI. The news reached the legal sector with the speed that this type of story usually does, and the most immediate conclusion was predictable: AI is dangerous, this case proves it. The problem is that the case doesn't prove that. What was sanctioned was the submission of citations of non-existent case law, generated by a language model and never verified before reaching real procedural documents. The duty of professional diligence did not originate with AI and is not affected by it either: any lawyer who cites case law without verifying it, regardless of the source, is failing in exactly the same way. AI made the failure more visible, faster, and more documentable, but the failure itself already existed as a possibility before any language model existed. When you heard about this story, what conclusion did you draw?