THE BOTTOM LINE
- A Closed AI Ecosystem for Courts: Judges are restricted to using only government-approved AI tools. This creates a controlled, predictable environment and a significant hurdle for unvetted legal tech providers hoping to enter the Spanish judicial market.
- Ultimate Judicial Accountability is Guaranteed: The new rules firmly place all responsibility for judicial decisions on the human judge, ensuring that AI-generated content is merely an assistive draft, not a final verdict. Businesses can be assured that final rulings will come from a person, not a machine.
- Strict Prohibitions on High-Risk AI Applications: The use of AI for profiling individuals, predicting behavior, or evaluating risk is explicitly banned. This reinforces data protection principles and limits AI’s role in sensitive judicial assessments, particularly in criminal and data-heavy civil cases.
THE DETAILS
Spain’s General Council of the Judiciary (CGPJ), the governing body for the country’s judges, has issued a landmark set of instructions on the use of artificial intelligence in judicial activities. The goal is to provide a clear and uniform framework that aligns with new Spanish and European regulations, including the EU AI Act. This move is not a ban but a carefully constructed set of guardrails, acknowledging the potential of AI to create efficiencies while proactively mitigating the risks of algorithmic bias and the erosion of fundamental rights. The instructions establish that any AI tool, especially generative AI, must be used in a way that respects judicial independence and ensures the integrity of the legal process.
The cornerstone of the new framework is the principle of effective human control. The CGPJ makes it unequivocally clear that AI systems are to be treated as support tools, not autonomous decision-makers. Judges and magistrates must maintain real, conscious, and effective control at all times. This means AI is explicitly prohibited from autonomously making judicial decisions, assessing facts or evidence, or interpreting and applying the law. The rules are designed to prevent a scenario where a judge is simply “replaced” or unduly influenced by an algorithm, ensuring the final legal reasoning and decision remains an exclusively human endeavor.
On a practical level, the instructions define what is and is not permissible. Judges may use approved AI tools for tasks like legal research, reviewing case precedents, and structuring information. They can even use AI to generate drafts of judicial resolutions, but only as a starting point. Every AI-generated draft requires a “complete and critical personal review and validation” by the judge, who remains solely responsible for the final text. Critically, the use of AI tools not provided or sanctioned by the justice administration or the CGPJ is forbidden. Furthermore, AI cannot be used for processing specially protected personal data or for any purpose related to profiling individuals, predicting their behavior, or classifying them based on risk assessments.
SOURCE
Consejo General del Poder Judicial (CGPJ)
