The OECD.AI Policy Navigator

Complex government documents may confuse citizens, making it difficult to access services such as pensions or benefits. This work uses GenAI to rewrite these texts in plain language, maintaining legal accuracy while increasing clarity. Focused on the Service Pages of Italy's National Institute of Social Security portal, which serves 45 million users, AI-generated versions received strong user preference over the original texts, enabling the Institute to communicate inclusively and at scale.

The purpose of the solution is to establish a system for the automatic sorting of incoming mail and documents using AI in IS KRPAN. In the initial phase, it will be used by the Ljubljana Administrative Unit and is expected to increase document-processing efficiency, reduce manual intervention, shorten mail-processing times, lower the risk of errors, increase employee productivity and optimise operating costs, thereby benefiting employees and citizens.

The Judicial Branch (PJMT) adopts an AI governance model structured through strategic and operational committees responsible for guiding, prioritising, and overseeing institutional initiatives. The AI Center serves as the central hub, integrating the development, management, and application of AI across the judiciary. The model includes feasibility analysis, approval workflows, and budgetary, ethical, and legal controls, aligned with National Council of Justice (CNJ) Resolution No. 615/2025.

Citizens send questions to INPS through a web form on different topics. In the past, all messages were checked by contact centre operators at the first level and manually forwarded where necessary to back-office officials at the second level. As a result, most requests were processed twice and answers arrived later. The new system uses AI to automatically send each request to the correct level. This helps citizens receive faster replies and reduces unnecessary work for contact centre staff.

The Ministry is implementing a cross-cutting initiative to integrate AI into public administration, focusing on human resource management and the simplification of administrative procedures. The initiative uses AI and data-driven tools to support workforce management, skills development and decision making, while redesigning and simplifying processes to reduce administrative burden. It aims to improve efficiency, transparency and the quality of services for citizens and businesses.

The Ministry of the Interior introduced an AI-enabled initiative to support a permanent system of incentives and rewards for public employees, based on key performance indicators (KPIs). The initiative uses AI-supported digital tools to help link individual and organisational performance with measurable results, while decisions remain under human oversight. It was developed to strengthen performance-based rewards and promote fairness, transparency and motivation across public administration.

This initiative provides guidance on how personal data should be handled throughout the lifecycle of generative AI systems. It is intended for organisations and individuals who develop, deploy or use generative AI systems that may process personal data. The guideline was developed to address privacy risks arising from the adoption of generative AI systems and to promote responsible and rights-respecting use of these technologies.

Mentor.AI is an artificial intelligence chatbot designed to support hiring agents in the planning stage of government procurement. The tool offers more agility and accuracy when consulting current and updated regulations and guidelines, as well as increasing the transparency and security of the processes.

This AI-driven initiative automates the identification of legal overlaps and contradictions within the National Registry of Legislation. It helps government officials and legal experts ensure that new regulations do not conflict with existing laws. By streamlining legislative drafting, the tool benefits legal workers by providing a simplified legal environment.

This initiative uses artificial intelligence to automatically summarise appeals submitted by citizens to the Social Security administration. It helps caseworkers quickly understand the main issues raised and classify each appeal for the appropriate handling. The system was developed to speed up processing times and manage growing workloads while ensuring citizens’ appeals are handled more efficiently.

The Government of Newfoundland and Labrador’s Responsible Use of Artificial Intelligence Policy sets clear rules for how AI can be used safely and responsibly across government. It ensures AI is approved, transparent, and subject to human oversight, while protecting privacy, fairness, and public trust. The policy supports responsible innovation and consistent decision-making across departments.

Bogotá, through its District Commission for Digital Transformation, created official guidelines to guide public entities in the responsible and transparent use of artificial intelligence. The initiative sets out clear rules and a common process with human oversight and risk management. It responds to the rapid growth in the use of AI and aims to build trust and harness innovation to strengthen Bogotá’s position as a digital capital.

A satellite-based AI system was developed for the public water management authority JP “Vode Vojvodine” to detect waterlogging and map risk areas across Vojvodina, helping them make faster and more informed decisions. The platform combines satellite images, weather data and soil information to identify where excess water may cause damage. It was created to reduce agricultural losses and support better planning, prevention and response to extreme weather conditions.

This initiative aims at the development and deployment of sovereign, state-operated AI assistants across public administration websites and portals to support citizens with information and public services. It addresses service accessibility and contact centre capacity by providing faster and more reliable digital assistance.

The initiative is a flexible module for automatic text correction, designed to improve efficiency, quality, and reliability in writing and communication. While targeted at “JP Službeni Glasnik”, it potentially benefits institutions, professionals, and anyone working with Serbian-language content. Developed to address the need for accurate OCR and text analysis, it opens opportunities for integration with existing tools and broader language applications.

VodostAI is a software platform that uses AI and data from IoT sensors to help predict floods and support early warnings. It helps water authorities, municipalities, and emergency services prepare and respond sooner, reducing damage and risk to people. It was developed to improve forecasting and decision-making where data are limited or arrive too late for effective action.

This initiative develops and strengthens Science and Technology Parks in Serbia as innovation infrastructure for AI development, providing specialised infrastructure, mentoring, and support services for AI start-ups and innovative companies. By fostering collaboration with research and higher education institutions, it accelerates AI-based entrepreneurship, supports commercialisation of AI solutions, and drives economic growth, employment, and the adoption of advanced technologies nationwide.

The initiative leverages artificial intelligence and big data analytics to strengthen risk based approaches to personal income tax compliance. By identifying atypical taxpayer behaviour, it supports more targeted audits and efficient allocation of enforcement resources. The project contributes to the broader modernisation of the Serbian Tax Administration and promotes evidence based policymaking in tax administration.

The initiative uses artificial intelligence to predict daily water inflows into reservoirs in the Drina River basin. It helps water managers and energy operators plan more effectively. It was developed to improve forecasting accuracy and support better decision making in water and energy management.

The AI Strategy lays out a plan for how the government will move forward with AI from 2025–2027. It highlights the importance of coordinating across federal departments, learning from collective experiences, and supporting departments of all sizes and maturity as they explore this technology. It also sets out the role of leaders in adapting AI to their organisations’ mandates. This role includes ensuring strong data foundations and the use of agile practices.