The OECD.AI Policy Navigator

Our policy navigator is a living repository from more than 80 jurisdictions and organisations. Use the filters to browse initiatives and find what you are looking for.

Trending Topics in Ombudsman Submissions


-
Added by:   OECD analyst
Added on:   28 Oct 2026
Updated by:   OECD analyst
Updated on:   28 Oct 2026

Trending Topics in Ombudsman Submissions uses artificial intelligence to identify common themes in submissions sent by citizens to the Brazilian Federal Government through the Fala.BR ombudsman platform. It shows how these topics change over time and highlights sudden increases in complaints or concerns. This helps public institutions respond faster and coordinate actions across agencies.

Name in original language

Identificação de Tópicos em Manifestações de Ouvidoria

Initiative overview

Trending Topics in Ombudsman Submissions was developed to address the difficulty of summarising, in clear and comprehensive themes, the content of ombudsman submissions received by Brazilian Federal Executive Branch entities through the Fala.BR platform. Although the platform includes structured classification fields—such as type, subject and sub-subject—these are initially filled in by citizens, who may not be familiar with government structures and therefore often classify their submissions inaccurately. In addition, these predefined categories are not detailed enough to fully capture the issues raised. The most relevant information is contained in an open-text field, written in free form by citizens with diverse educational backgrounds and language styles.

Given the high number of submissions received, it is not feasible for small ombudsman teams to manually produce aggregated and meaningful analyses that accurately reflect public concerns and their intensity. At the same time, the centralisation of all Federal Executive Branch ombudsman submissions within the Fala.BR platform—maintained and managed by the Office of the Comptroller General (CGU)—creates a unique opportunity for centralised access and large-scale analytical processing of this citizen feedback.

The primary objective of Trending Topics is to identify common themes within the textual content of ombudsman submissions and to monitor how these themes evolve over time, including detecting sudden increases in related complaints or concerns. The initiative aims to strengthen the State’s capacity to listen to society in a structured way, to improve coordination and timeliness in government responses, and ultimately to enhance public trust in government institutions.

Currently, the solution is used by the Office of the Ombudsman General (OGU) within the CGU (1) to produce periodic analytical reports for federal ombudsman units, (2) for its own supervisory activities, and (3) for senior leadership to monitor strategic public policies and services. In the future, the tool will be integrated directly into the Fala.BR platform, enabling individual ombudsman units to autonomously access and use the analytical functionality.

The initiative is institutionally supported by Brazil’s General Data Protection Law (Law No. 13.709/2018), the Public Service User Protection Law (Law No. 13.460/2017), and Decree No. 9.492/2018, which established the Federal Executive Branch Ombudsman System.

Other relevant details

Results, outcomes and impacts: The initiative has improved monitoring of federal ombudsman units by identifying recurring themes and explaining sudden peaks in complaints. More than 150 analyses have been conducted, and management reports have been shared with around 20 ministry-level ombudsman offices. Insights generated by the tool have informed senior leadership at CGU about relevant issues in public policies and services, leading to formal engagement with two federal entities. In the future, it is expected to enable faster and more co-ordinated corrective actions by government agencies, strengthening public trust. Challenges and lessons learned: A key challenge was handling sensitive and legally protected data. AI use was therefore designed to avoid public services, relying instead on on-premises models or contracted solutions with strong data protection safeguards. Another challenge was the lack of clear metrics to assess topic quality, creating risks of inaccuracy or large language model (LLM) hallucinations. This has been mitigated through careful human review before sharing results. Reports are communicated cautiously, highlighting the pilot nature of the initiative to reduce reputational risks. A key lesson is the importance of understanding AI limitations and embedding human oversight. Clear governance, defined workflows, and adequate technical and human resources are essential for success.

About the policy initiative


Category:

  • AI policy initiatives, programmes and projects

Initiative type:

  • AI use cases/projects in the public sector

Status:

  • Active

Start Year:

  • 2025

Target Sectors:


OECD AI Principles:

—

Other relevant urls:

—