The Bottom Line
- Restricted Tech Ecosystem: Spanish judges are now limited to using only government-approved and CGPJ-vetted AI tools, creating a controlled market for legal tech providers targeting the judiciary and signaling a significant risk in using unapproved public AI for legal work.
- Unaltered Judicial Responsibility: AI cannot be used to automate final judicial decisions, evaluate evidence, or interpret law. The principle of effective human control is paramount, meaning CEOs and General Counsel can be assured that a human judge remains fully and personally responsible for every ruling.
- Strict Data & Profiling Ban: The guidelines explicitly prohibit using AI for processing specially protected personal data or for profiling individuals, predicting behavior, or conducting risk assessments. This reinforces high data security standards and limits the scope of AI application in sensitive legal contexts.
The Details
Spain’s General Council of the Judiciary (CGPJ), the governing body of the country’s judiciary, has issued a landmark instruction that sets clear and firm boundaries for the use of artificial intelligence by judges and magistrates. This proactive measure aims to provide a coherent framework that aligns with new EU AI regulations, ensuring that the adoption of technology does not compromise judicial independence or fundamental rights. The instruction recognizes the potential of AI as a support tool while decisively addressing the risks posed by unchecked automation, particularly from generative AI systems.
The core principle underpinning the new rules is effective human control. The CGPJ mandates that judges must maintain constant, conscious, and effective oversight of any AI system they use. The instruction is unequivocal: AI is an assistant, not a substitute. It explicitly prohibits the automation of judicial decision-making, the assessment of facts or evidence, and the interpretation or application of law. While judges can use approved AI to help draft internal working documents or summaries, they bear exclusive responsibility for the final text and must conduct a “complete and critical personal review” before incorporating any AI-generated content into an official resolution.
The guidelines establish a clear distinction between permitted and prohibited uses, effectively creating a walled garden for judicial AI. Judges may only use applications provided and vetted by the competent justice administrations or the CGPJ itself. Permitted uses include legal research, analysis of procedural history, and structuring information to aid in case analysis. Conversely, the instruction strictly forbids using AI tools for profiling people, predicting their behavior, or classifying them based on risk. Furthermore, any processing of specially protected personal data with AI is banned, ensuring that the judiciary maintains the highest standards of confidentiality and security.
Source
Consejo General del Poder Judicial (CGPJ)
