Maldives offers free AI prosecution governance framework: where do the limits of the machine lie in courts

The Maldives announced a unusual regulatory step by opening its framework for AI governance in criminal prosecution and its training methods to all countries and judicial bodies worldwide, free of charge. The announcement was made by Maldives Attorney General Abbas Sharif during his participation in the United Nations Conference on Crime Prevention and Criminal Justice held in Abu Dhabi, the UAE capital, in an effort to provide a practical model that integrates algorithms into the judicial system while strictly keeping charging decisions in the hands of humans alone.

The Maldives framework rests on a decisive rule that prevents the machine from deciding to initiate a criminal case or refer any person to trial, leaving that authority confined to prosecutors and judges subject to institutional oversight. The system also imposes a mandatory transparency requirement, obliging prosecution teams to disclose publicly and officially whenever artificial-intelligence tools are used in drafting or preparing memoranda and documents submitted to the judicial platform, a measure intended to preserve legal accountability and prevent the infiltration of biases or algorithmic errors without announced human review.

The machine may assist a prosecutor but it will not replace human judgment.That is how the Maldives Attorney General summed up the equation that falls within the expanded government digital transformation program known as Maldives 2.0. This approach coincided with acute real-world pressures on law-enforcement agencies there, after cyber-crime cases doubled in the first half of 2026 amid complex difficulties in tracking cross-border financial transactions, gathering sufficient digital evidence and pursuing suspects abroad, which prompted the state to amend the penal code, the criminal procedure code and international legal-aid legislation in line with the Budapest Convention on cybercrime.

This open move has a direct and practical impact on justice systems and legal technology in the Arab world, particularly in the Gulf, Egypt and Jordan. Justice ministries and public prosecutors that spend massive advisory budgets to develop evidence of generative AI use now have a ready-made, free procedural standard they can rely on instead of starting from scratch, reducing regulatory drafting costs and establishing a mandatory ceiling for transparency standards; requiring the prosecution to disclose before the court any content generated by artificial intelligence gives defense lawyers and oversight bodies a new auditing tool against digital conviction reports.

The model also requires justice-software developers in the region to recalibrate their engineering tools so that they stop entirely at the limits of summarising, sorting and gathering facts, excluding any attempts to automate legal adaptation or suggest rulings, in line with the mandatory human-oversight requirements adopted by emerging international legislative frameworks to protect individuals’ rights and fundamental freedoms.