Strengthened control over judicial data used for AI

Lawyers join forces with magistrates on the impact of artificial intelligence systems on the administration of justice. The OCF (Congressional Body of Lawyers) thus aligns itself with the priorities expressed by the judiciary, summarised in a document signed by over 50 senior officials from judicial offices (including the Anti-Mafia Prosecutor, Giovanni Melillo, the President of the Milan Court of Appeal, Giovanni Ondei, and the Attorney General of Rome, Giuseppe Amato) and suggests shared governance.

A number of key points, such as data sovereignty in the justice system: court documents, wiretaps, and information on minors and victims, must be processed on public or certified infrastructure and must never be reused to train commercial models; the commercial confidentiality that the implementing decrees of the enabling act on this matter extend to algorithms and training data must not be allowed to be invoked against public scrutiny and the right to a fair trial.

Oversight cannot be entrusted solely to those who develop, supply or use artificial intelligence systems. We need independent third-party verification and audit bodies, drawing on legal, technological and scientific expertise.

Ocf emphasises that there must be no ‘black boxes’ in the process: if artificial intelligence is used in judicial proceedings, its use must be transparent and verifiable by the defence.

Equality of arms is also crucial: innovation must not lead to a two-tier justice system in which the judiciary, the prosecution and the defence have differing levels of access to data, technological tools and knowledge of how they work.