The Minister of Science and Technology discusses AI in elections during a project meeting

The seventh cycle of debates of the Mídia e Democracia Council featured the participation of the Minister of Science, Technology, and Innovations, Luciana Santos. The meeting, themed "Artificial Intelligence, Disinformation, and Elections – Possible Uses in Electoral Campaigns: Opportunities and Risks in Expanding Access and Regulating This Technology," also addressed disinformation and included a workshop conducted by Democracy Reporting International (DRI) and Lupa.

During the plenary session, the minister presented the government’s initiatives to regulate Artificial Intelligence in Brazil through the "AI for the Good of All" plan, which focuses on developing and adopting an advanced AI model in Portuguese. Investments total R$25 million, aiming to strengthen the technology and position Brazil as a global leader in the field. This plan seeks to establish an adequate structure by leveraging expertise from public universities and state-owned companies, such as Dataprev and Serpro.

Key points highlighted by council members during the discussions included how the government plans to regulate AI by considering thematic chambers and aligning state-level initiatives. Ensuring transparency regarding the resources allocated within the AI plan was another significant concern. Both the minister and council members emphasized the importance of developing domestic platforms and low-orbit satellites for telecommunications to reduce dependency on foreign companies.

The workshop titled "Artificial Intelligence and Disinformation: How AI Can Be Used for Disinformation, Deepfakes, and the Difference Between Synthetic and Manipulated Content" aimed to align concepts and identify AI tools utilized during elections while reflecting on governmental regulation. The discussion outlined the pillars of disinformation, divided into three categories: tools - the channel and technical foundation of the message -, tactics - how the message is distributed -, and narratives - the content of the message that leverages the other two.

To illustrate how the two organizations are addressing AI, DRI shared a guide developed by the organization on AI tools, showcasing their use in electoral contexts worldwide, including India, Argentina, and the United States, as well as cases in Brazil, particularly during the environmental disaster in Rio Grande do Sul. Lupa, on the other hand, presented a series of fact-checks where AI was used as a tool.