Today's Edition
ABU DHABI (ALETIHAD)
The Ministry of Justice is accelerating the digital transformation of the judicial system through the integration of agentic artificial intelligence, reducing the time required to review case files from several days to just a few hours while retaining full judicial oversight and decision-making by judges.
The initiative is being showcased during the Ministry's participation in the 15th United Nations Congress on Crime Prevention and Criminal Justice - Abu Dhabi 2026, hosted by the UAE at ADNEC Centre Abu Dhabi until October 1.
The Ministry is presenting a suite of specialised AI agents designed to support judges throughout different stages of the judicial process, demonstrating how advanced technologies can improve the efficiency and quality of judicial services.
Five specialised AI agents
Visitors to the Ministry's pavilion at the exhibition accompanying the Congress can follow the digital journey of a case, from the submission of a digital case file through to the preparation of an initial draft judgment.
The showcase features five specialised AI agents integrated with the Ministry's existing electronic court systems: the Case File Review Assistant, Case Analysis Assistant, Legal Research Assistant, Decision Support Assistant and Judgment Drafting Assistant.
The Case File Review Assistant classifies and processes digital case documents and extracts key information, while the Case Analysis Assistant links facts and evidence and identifies the principal legal issues.
The Legal Research Assistant compiles relevant legislation and judicial precedents. The Decision Support Assistant summarises the legal options available to the judge and their implications, while the Judgment Drafting Assistant prepares a non-binding preliminary draft judgment.
The draft remains fully subject to the judge's review, amendment and approval, ensuring that final judicial decisions remain under human authority.
The five agents operate in a coordinated sequence.
Human oversight
His Excellency Abdulrahman Murad Al Blooshi, Assistant Undersecretary for the International Cooperation and Legal Affairs Sector at the Ministry of Justice and General Coordinator of the 15th United Nations Congress on Crime Prevention and Criminal Justice - Abu Dhabi 2026, said the model combines artificial intelligence with human oversight across a seven-stage review process.
"The model is built on an integrated framework that combines artificial intelligence with human oversight. The review process consists of seven stages, five of which are supported by AI agents, while the remaining two stages remain exclusively under the responsibility of judges.
"This approach ensures that judicial discretion and final decision-making remain entirely in the hands of the judge," he said.
The distribution of responsibilities between AI systems and judges is designed not only to accelerate judicial procedures, but also to reduce routine preparatory work, allowing judges to focus more closely on the substance of cases and the assessment of facts.
According to the Ministry, this balance is intended to contribute to both the quality of judicial decisions and the speed of proceedings, while establishing human oversight as a fundamental element of digital transformation in the justice sector.
Al Blooshi added: "This initiative reflects the Ministry of Justice's leadership in promoting the responsible use of artificial intelligence and its commitment to enhancing judicial efficiency, accelerating the resolution of cases and delivering timely justice."
Through the deployment of AI-powered judicial assistants, the Ministry is working towards a justice system that is more efficient, responsive and user-centred, while maintaining the legal safeguards and human oversight that underpin the administration of justice.
The Ministry said the initiative represents an important milestone in its ongoing digital transformation journey, with the deployment of AI agents set to expand progressively following further evaluation, testing and integration with judicial partner entities.
