Source: The Kathmandu Post

Kathmandu, 3 July 2026: The Supreme Court’s full-court meeting endorsed the Judiciary Information and Communication Technology Policy 2026, authorising AI use in courts for legal research, document management and administrative support while expressly prohibiting AI from writing judgments, writ petitions or other outputs requiring judicial discretion. The policy allows AI to draft non-decisional documents and support academic study but bars any AI influence on judicial decision-making and stresses the need for human legal reasoning in dispositive acts. Any AI system—whether developed internally or by outside organisations—must be assessed for reliability, transparency, potential bias and its operational framework before it can be used in judicial work, with officials warning that AI could generate incorrect or non-existent legal references.

The judiciary will restructure its case management database to store records in machine-readable formats so AI can rapidly search and analyse Supreme Court and lower courts’ judgments and orders. Clause 5.2 sets a staged rollout of AI as an innovation tool, coupled with awareness programmes and training for judges, court officials, lawyers and court users to address benefits and risks. Supreme Court Chief Registrar Bimal Paudel and joint secretary Amit Upreti stressed safeguards to ensure AI assists rather than replaces human decision-making. The court plans AI assistance for organising and managing court orders, training for the Nepal Bar Association and government attorneys, and may seek partnerships with government agencies, the private sector, academic institutions and international development partners to strengthen digital infrastructure; Chief Justice Manoj Kumar Sharma said the policy fulfils his e-courts pledge and will be updated over time.