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.

ANAC Artificial Intelligence Strategy 2025–2026


Added by:   OECD analyst
Added on:   03 Aug 2026
Updated by:   OECD analyst
Updated on:   13 Oct 2026

ANAC’s AI Strategy 2025–2026 is the Brazilian Civil Aviation Agency’s roadmap for adopting AI across the organisation in a broad, safe and responsible way. It links governance, capacity-building, risk management and implementation. It was created to improve public services, strengthen decision support and protect data, rights and trust.

Name in original language

Estratégia de Inteligência Artificial da ANAC 2025-2026

Initiative overview

ANAC’s AI Strategy 2025–2026 is the formal roadmap that operationalizes the agency’s AI Policy and translates AI adoption into concrete goals, governance arrangements, actions and monitoring routines. The strategy responds to a clear opportunity and challenge: to use AI to improve internal work, digital services and regulatory performance while ensuring safety, ethics, transparency, data protection and institutional accountability. The initiative is closely aligned with ANAC’s broader institutional planning. It connects to the agency’s Strategic Plan 2020–2026, its Digital Strategy 2024–2026 and the federal digital government agenda. Its strategic vision is to achieve excellence in the broad, safe and responsible adoption of AI to advance ANAC’s mission. To support this, the strategy defines key results for staff adoption and service transformation, and organizes implementation around five objectives: integrating AI into staff activities, increasing the capacity to develop AI-enabled solutions, improving data governance, ensuring security and ethics, and optimizing AI costs. The strategy is also a governance mechanism in practice, because it assigns roles and responsibilities across the institution. ANAC’s normative framework establishes an AI governance structure involving the IT Committee, the Superintendence of Technology and Digital Transformation, the data protection officer, technical advisory units, director-linked units and AI focal points. It also created an AI core with representatives from nearly all units, who met weekly, received training and helped design the policy and strategy. This creates a cross-functional model for prioritization, oversight, risk handling, implementation support and organizational learning. Looking ahead, the initiative is designed to become more institutionalized and scalable. ANAC established a roadmap, an operational model, periodic monitoring and evaluation routines, and requirements for existing AI systems to adapt to the new rules. The strategy foresees broader uptake across internal and external services, additional specific procurements if needed, and stronger reuse of models, standards and knowledge across the agency. In this sense, the initiative is not only a strategy document, but an institutional framework for trustworthy AI adoption in a federal regulator.

Other relevant details

Results, outcomes and impacts: KAŞİF delivered immediate results by identifying high-risk fake invoicing activities as they emerged. The system detected newly established taxpayers issuing high-value invoices without real economic activity and uncovered coordinated networks of shell companies across various tax offices. By flagging thousands of suspicious entities and blocking unjustified VAT refund claims, KAŞİF effectively prevented significant public loss and ensured the integrity of the tax system. Challenges and lessons learned: Key challenges included integrating heterogeneous data sources, ensuring data quality at scale, and aligning AI outputs with existing legal and operational processes. These were addressed through strong inter-institutional collaboration, continuous data validation, and close integration between analytics teams and tax inspectors. An important lesson learned is that AI systems must complement, not replace, expert judgement. Clear governance, secure national infrastructure and ongoing model monitoring are critical conditions for the success of large-scale AI initiatives in government.

About the policy initiative


Category:

  • AI policy initiatives, programmes and projects

Initiative type:

  • Other AI policy initiatives, programmes and projects

Status:

  • Active

Start Year:

  • 2025