CCBE Guide: Ethical Liability in the Use of Generative AI

1. The facts

   The issue of lawyers' liability for AI use gained global relevance following a well-known international case, extensively documented by outlets like Il Sole 24 Ore and Diritto e Giustizia. In that instance, a lawyer submitted a brief containing six non-existent judgments generated by ChatGPT. In response, at the European and national levels, action was taken through the CCBE guide and the October 2025 informative model. At present, the matter is in a phase of soft-law production and advisory orientation, aimed at preventing disciplinary proceedings through the correct interpretation of existing rules on professional diligence. Other aspects of the story, related to lead generation and the reform of the legal profession, are covered in dedicated articles in this column.

2. 

   The legal framework

   The regulatory framework centers on the Forensic Deontological Code and the Civil Code.

   1. Art. 26 CDF: imposes the duty of diligence, stating that lawyers must perform professional activities with accuracy, making legal research not fully delegable to automation.
   2. Art. 12 CDF: establishes the duty of competence, requiring the lawyer to use only tools whose functioning and technical limits they understand.
   3. Art. 1176 c.c.: defines professional diligence as a standard higher than ordinary diligence, making AI-generated errors potential gross negligence.
   4. Art. 9 CDF: protects the duties of probity and decorum, breached if documents containing falsehoods, even if software-generated, are presented to the judge.

3. Case law orientation

   Although specific decisions on algorithmic hallucinations are lacking, a doctrinal thesis has suggested that a lawyer is always responsible for the work of their assistants, a category into which AI would fall. According to this theoretical approach, the principle of liability for failure to verify would apply, whereby signing a document implies ownership of all content, ruling out that an error by a collaborator (or machine) can exempt the professional. Legal scholars emphasize that the duty of control is an essential element of the professional munus, and failure to check official sources constitutes a disciplinary violation.

5. Analysis drafted and verified with edit.legal

   To verify the provisions cited in this article, we used edit.legal. Test our legal AI on official sources and apply it to your own matters.

   Try edit.legal AI

8. Lessons for professionals

   1. Always verify every citation produced by AI by consulting exclusively official sources or certified databases.
   2. Maintain constant control of the decision-making process by applying the human-in-the-loop principle to every draft.
   3. Inform the client about the technological limits of the tools used to avoid unfounded expectations regarding research speed.
   4. Protect professional secrecy by avoiding the entry of sensitive client data into the prompts of unprotected AI systems.

References: Guida CCBEModello informativa ottobre 2025Codice Deontologico Forense, artt. 9, 12, 14, 26Codice Civile, artt. 1176, 2236

Editorial oversight: Avv. Federico Papa·ICAMUpdated 17 September 2026Content drafted with AI support and subject to editorial source checks. Despite these controls, inaccuracies may remain: reports and rectification requests are welcome. Report a correction