English
Quebec’s judiciary has drawn a firm line on the use of artificial intelligence in the courtroom, issuing new guidelines that bar judges from using generative AI tools for legal reasoning, evidence assessment or decision-making. The rules, adopted jointly by four of the province’s courts, make clear that judging must remain an exclusively human function.
Four Courts Unite on AI Restrictions
The Quebec Court of Appeal, Superior Court, Court of Quebec and the province’s municipal courts jointly adopted the 10-page policy document. It permits judges to use generative AI for limited administrative and editorial tasks, but draws a hard boundary around anything that touches judicial reasoning or deliberation.
Stay connected to every major update — subscribe and follow us on the PhoenixQ website and across our social media platforms.
“Generative AI, however sophisticated, possesses neither judgment nor conscience and is not subject to any form of accountability,” the guidelines state. “It cannot understand the human, social and legal context specific to each case.”
The courts were equally direct about the nature of judging itself. “Judging cannot be reduced to a technical exercise,” the document says. “It is an intrinsically human function, based on deliberation, consideration of context, respect for dignity, the requirement of fairness and the exercise of moral responsibility.”
A Controversy That Prompted Urgent Action
The guidelines follow a controversy that raised serious concerns within Quebec’s legal community. In March, La Presse reported that a judicial decision contained references to case law that did not exist. The report prompted speculation that a judge had used generative AI in preparing the ruling.
That episode created a sense of urgency inside the judiciary. Courts moved to establish formal rules before the use of such tools could spread without oversight or accountability.
The guidelines also note a practical concern: no approved, secure institutional AI tool currently exists for judges to use. Existing commercial tools rely on algorithms whose inner workings remain largely opaque, the courts warned.
What Judges Can and Cannot Do
Under the new rules, judges may use generative AI for a narrow set of tasks. These include correcting, revising, rephrasing or translating text they have already written, handling certain administrative work, and locating information within specified sources.
Any such use must remain under full human control. Judges must independently verify any AI-generated material before relying on it.
The restrictions go further when it comes to substantive legal work. Judges cannot use AI to analyze or characterize facts, assess evidence or credibility, determine the outcome of a dispute, or draft the reasoning sections of a decision.
“If the use of AI has the effect or purpose of determining the legal outcome, assessing credibility or interpreting disputed facts, interpreting the relevance of facts, or formulating the reasons for a decision, it must not be used,” the guidelines state.
Confidentiality and Full Judicial Accountability
The courts also raised confidentiality concerns. Judges must not feed AI tools with draft judgments, notes about ongoing cases, or information protected by law or court order.
Until an authorized institutional system becomes available, judges using AI for permitted tasks must ensure the tool’s settings prevent any information they enter from being used to train or improve the system.
Crucially, the guidelines make clear that no error can be blamed on the technology. “The use of a generative AI tool does not diminish the judge’s obligation to verify or their full responsibility for the final result,” the document states. Judges remain fully accountable for their legal reasoning, the accuracy of facts, citations, quotations and legal authorities in every decision they issue.
The courts say the guidelines are designed to protect judicial independence, impartiality, fairness, human rights and access to justice, while also safeguarding confidentiality and data security. When any doubt arises about whether a particular use is appropriate, the guidelines advise judges to refrain from using generative AI altogether.
