AI and Government Audit: Reflections by Silvio Meira at TCU seminar

The morning of October 22 was marked by a lecture from Silvio Meira, scientist, professor, and one of the most influential voices in the country on innovation and digital transformation. Speaking at the International Seminar The Future of Public Auditing: Data, Innovation, and Citizenship , promoted by the Brazilian Federal Court of Accounts (TCU) and the Serzedello Corrêa Capacity Development Institute (ISC) in Brasília, Brazil, he offered a thought-provoking reflection on the role of artificial intelligence in the oversight environment, highlighting how technology can transform governance and auditing processes in public administration.

Meira began by contextualizing the impact of AI on society and markets, stating that the world is undergoing another major revolution. According to him, AI is no longer merely a tool but has become a true immersion in human intelligence, capable of processing information faster and more efficiently than humans. He explained that AI operates at multiple levels of intelligence, from descriptive (analyzing what is happening now) to strategic (defining, executing, and adapting strategies according to context).

In the context of public oversight, Meira pointed out that AI can be a powerful ally in increasing transparency, efficiency, and agility in audit processes. The explanation is simple: AI enables the analysis of large volumes of data in real time, helping civil servants identify patterns, anomalies, and potential irregularities. However, governance over this data is essential.

"Studies show that businesses that fail to achieve prescriptive intelligence* level by 2030 will face monumental challenges ahead. But to do that, we need data, and data is not a trivial matter of having or not having. All business captures and inputs data in some way. The data must be useful for decision-making based on the reality it reveals. And, obviously, security and privacy are something we need to comply with by legal requirement. Very few organizations truly manage their information lifecycle strategically both within and for the business - to be used with artificial intelligence in decision-making," points out Meira.

For the professor, innovation is essential if audit institutions are to keep pace with changes in markets and society. Still, he made it clear that the future of public oversight involves integrating artificial intelligence and adopting adaptive governance practices. Both crucial to building an effective public oversight model grounded in integrity and transparency. We need adaptive architectures that shorten decision cycles, with ethical safeguards and controlled testing to learn from mistakes and continuously evolve, he suggested.

During the lecture, Meira also referred to his latest book, A Próxima Democracia (The Next Democracy), co-authored with Rosário Pompéia. The book presents 32 theses for reestablishing politics in the phygital world (integration of physical, digital, and social dimensions, creating a dimension where these three spheres intertwine and coexist in an interdependent manner). He pointed out that many of these theses are applicable to the oversight environment, especially those concerning transparency, data governance, and adaptive innovation.

About the International Seminar The Future of Public Auditing

In its first edition, the seminar featured both theoretical presentations and practical demonstrations on four central themes: Citizen-Centered Public Auditing; Skills and Mindset of the Future Auditor; Innovative Approaches in Public Sector Auditing: Cases and Lessons Learned; Artificial Intelligence and Digital Transformation in Public Auditing.

A total of 73 papers selected by the evaluation committee were presented, 27 as lectures and 46 as posters. The presentations showcased innovative oversight solutions and experiences, with contributions from both national and international experts.

*Note: Prescriptive intelligence is the level of artificial intelligence that analyzes what is happening (descriptive intelligence), predicts what is likely to happen (predictive intelligence), and indicates what should be done to achieve a certain goal or solve a problem.

Learn more at: International Seminar | The Future of Public Auditing