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.

Official Gazette Intelligence (Subnational: Public Management Innovation Consortium)


Added by:   OECD analyst
Added on:   12 Nov 2026
Updated by:   OECD analyst
Updated on:   12 Nov 2026

undefined
CIGA Diário Intelligence is an AI-driven platform for Brazilian municipal official gazettes. It automates data privacy compliance (LGPD), generates smart document summaries for citizens, and interprets legislative impacts by cross-referencing new and existing laws. It also ensures system integrity by flagging malicious content for human review. By acting as a co-pilot for public managers, the tool enhances transparency and legal security while maintaining strict human oversight.

Name in original language

Ciga Diário Intelligence

Initiative overview

The Public Management Innovation Consortium (CIGA) operates the Municipal Official Gazette (Ciga Diário), a critical transparency tool for over 295 Brazilian municipalities. With more than 100,000 documents published annually by 4,500 different users, manual oversight of every publication was historically impossible. This high volume, combined with the stringent requirements of the Brazilian General Data Protection Law (LGPD) — which is closely modelled after the European Union's GDPR — created a significant risk of privacy breaches and administrative errors. CIGA Diário Intelligence was developed to bridge this gap, leveraging AI to provide a layer of security and intelligence that ensures transparency does not come at the cost of data privacy or legal integrity.

Although the Gazette system has successfully operated for over a decade, the integration of artificial intelligence in 2025 marked a transformative evolution. The initiative was catalysed by the need for regulatory compliance, a pro-innovation internal culture, and the increasing accessibility of high-quality AI models. The platform now operates across four strategic pillars: automated privacy analysis to detect sensitive data, AI-generated summaries to improve citizen accessibility, intelligent legislative compilation to track legal impacts, and content integrity monitoring. This shift from a simple PDF repository to an intelligent governance hub allows public managers to handle complex documentation with greater precision.

The impact of the initiative is already measurable, particularly in risk mitigation. Previously, responsibility for content accuracy rested solely on individual users, which occasionally led to the publication of incorrect or sensitive information. Today, AI provides real-time moderation. In a recent high-stakes incident, the system demonstrated its value by identifying and blocking malicious publications attempted by a hacker who had compromised user credentials. By acting as a proactive guardian rather than a reactive archive, the tool prevents legal and reputational damage to the municipalities, proving that AI-assisted oversight is essential for modern public administration.

Looking ahead, CIGA is focused on the full institutionalisation of these features. While malicious content detection is already active for all users, the remaining functionalities are being phased in, with the goal of 100% adoption by the end of 2026. The current strategic focus is the continuous refinement of these models to ensure maximum reliability and accuracy. By providing this technology as a shared service through a public consortium, CIGA demonstrates a scalable and sustainable model for how small and medium-sized municipalities can access cutting-edge AI tools to improve local governance without the need for individual high-cost investments.

Other relevant details

Results, outcomes and impacts: So far, the AI has blocked malicious publication attempts from compromised accounts and led to a decline in support tickets for sensitive data removal. Success is monitored via system logs, user feedback, and help desk analytics. This adds a critical security layer to 100,000 annual publications that previously lacked centralised moderation, shifting the focus from reactive corrections to proactive prevention. By 2026, we expect to reduce sensitive data exposure incidents by 90%, ensuring 100% LGPD compliance for over 295 municipalities. Our goal is to become Brazil's national benchmark for automated legislative compilation and data privacy in public administration. Challenges and lessons learned: The primary challenge was integrating AI into a decade-old legacy system while ensuring high accuracy for LGPD compliance. With 4,500 users, the risk of data breaches or malicious content was significant. We responded by adopting a "Human-in-the-Loop" model, where AI acts as a co-pilot, flagging anomalies for human review rather than making autonomous legal decisions. A key lesson is that AI in government must augment, not replace, human judgement to maintain public trust. We also learned that security is paramount; our models successfully blocked a real-world attempt to publish malicious content via compromised credentials. Success requires a robust institutional culture of innovation, a clear regulatory framework (like LGPD/GDPR), and a shared-service model that enables small municipalities to access advanced technology. Ultimately, the condition for success is balancing technological efficiency with the strict accountability required in public administration.

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

OECD AI Principles:

—

Other relevant urls:

—

Images: