FGV hosts German Vice-Minister for Education and Research to discuss information integrity on Artificial Intelligence

On the morning of June 17, FGV hosted the Vice-Minister of the German Federal Ministry of Education and Research, Jens Brandenburg, to discuss the challenges of information integrity on artificial intelligence (AI). Accompanying the Vice-Minister were Susan Schulz, Deputy Head of the Division Cooperation with Latin America; Jan-Christoph Rogge, Personal Assistant to the Secretary; Joachim Schemel, Deputy General Consul of Germany in Rio de Janeiro; and Nina Sartori, Counsellor for Scientific Affairs at the German Embassy. Mr. Brandenburg commented on the impact of disinformation and misinformation on education in democracies, and also on artificial intelligence and responsible research, together with directors and researchers from FGV, as well as members of the Ethics Compliance Committee for Research Involving Humans and FGV's Central Office for Research Integrity.

For the German Vice-Minister for Education and Research, it is essential to educate the population about the use of AI. Mr. Brandenburg believes that it is necessary to focus on disclosure methods and rules for the use of information and data, as there is a high demand for the use of information related to citizens, but not always an adequate way of obtaining it. "The exchange of experiences between Brazil and Germany on this subject will boost collaboration between the two countries," said the Vice-Minister.

(Vice-Minister of the German Federal Ministry of Education and Research, Jens Brandenburg)

The Vice-Director of the School of Communication, Media and Information (FGV ECMI), Amaro Grassi was the moderator on the first panel debate, which focused on how disinformation and misinformation affect educational systems and democratic values. Grassi began the debate by highlighting the partnership with the German Embassy, through the Digital Democracy research project.

Mr. Brandenburg emphasized that "this issue (use of Artificial Intelligence) should not be answered by the government alone, but rather in a transdisciplinary way with scientists". In addition, the Vice-Minister argued that this is a challenge that must be solved beyond the borders of a single country, reiterating the possibility of countries exchanging knowledge and experiences on the impact and political challenges of using artificial intelligence. "Brazil and Germany have important points and similarities when it comes to the challenges facing the use of AI," he said.

Luca Belli, Professor and Coordinator of the Center for Technology and Society at the Rio de Janeiro Law School (FGV Direito Rio), also took part in the debate. Belli is developing the CyberBRICS research project that seeks to map AI regulation and industrial policy in the BRICS, as well as identify the best practices and develop policy suggestions in cybersecurity and personal data regulation. For the researcher, it is important to emphasize how complex the issue of using AI is, especially regarding cybersecurity and how digital content can be artificially created and spread by exploiting AI systems.

"Recently the National Cybersecurity Committee was created, of which I am a member. What we can see in recent years is the democratization of the use of AI, which brings great benefits to support scientific research, but with side effects in relation to the increase in potential targets and perpetrators of cyberattacks, the complexity and scale of such attacks and the ease of creating and disseminating extremely convincing false information, such as the so-called 'deepfakes'," Belli warned.

Regarding the risks of using AI, Eduardo Mendes, a researcher at the São Paulo School of Economics (FGV EESP), adds that as well as the population needing to know how these tools work, they need to understand that each section of the population will discuss the use of AI in a different way. "As well as knowing how to use them, the population needs to know how these tools work, so that they can use them correctly and responsibly."

The researcher also argues that it is essential to use AI tools themselves to combat disinformation and the proliferation of misinformation, as the population is consuming information and news on a scale never seen before. "An AI tool to check the veracity of information and news could be an example," Mendes suggested.

Delving deeper into the risks, Vivianne Ferreira, Professor at FGV Direito SP - São Paulo Law school and Coordinator of the Ethics Committee for Research with Human Beings, pointed out that AI tools can speed up research activities, such as transcribing interviews, but it is necessary to consider that the research participants have given their consent for exclusive use of their data in a specific research project. 

However, the Researcher points out that transcribed interviews may contain personal data, and if this transcription is carried out by an AI tool, it is possible that this tool will use the personal data to improve itself. Ferreira therefore wonders how to unite these two aspects: the dignity of the participant and the use of tools capable of following ethical guidelines. She also warns of the risk of using AI for diversity criteria. 

"We have a Diversity Policy at FGV, and speaking for the school where I teach, one of FGV Direito SP - São Paulo Law school's main concerns is diversity in research. For this diversity to occur, it is necessary to take into account diverse points of view, and when using any AI tool for writing or analysis, we run the risk of the tool reproducing general points of view that ignore diversity," the researcher claimed.

In this sense, the Director of the School of Applied Mathematics (FGV EMAp), Cesar Camacho, adds that, in the context of regulating the use of AI, Brazil has yet to advance in relation to other countries, which is why it is necessary to think about how to convince governments to create this regulation: "FGV could play a fundamental role in proposing this regulation," said Camacho, who also believes that the certification of Data Science professionals is extremely important.

Regulating the Use of Artificial Intelligence 

In the second panel of the event, moderated by Alexandre Pacheco, Professor at the São Paulo Law School (FGV Direito SP) and Coordinator of the Center for Teaching and Research in Innovation, the discussion covered the benefits and restrictions of using AI in academic research activities. 

On the occasion, Pacheco made a comparison between the General Data Protection Law (LGPD) in Brazil and the General Data Protection Regulation (GDPR). For the professor, although the LGPD has been greatly influenced by the GDPR, it is necessary to take into account the high costs of implementing and complying with European-style standards for artificial intelligence in the research area: "We have already experienced this with the implementation of the personal data protection regime in Brazil and the projection for artificial intelligence is that these costs will be even higher."

In this sense, Juliana de Palma, coordinator of FGV's Central Office for Scientific Integrity, believes that one possible way of dealing with these technologies is to create regulation at different levels, with an exchange of information and experience between universities.

"We are in this period of transition in which we need to create a system of regulation and we don't need to wait for the government to create this regulation. Given the experience of researchers, we can collect this information and try to provide guidance and guidelines on the use of AI, even if they are only temporary, in order to foster scientific integrity at different universities," she said. 

The event continued throughout the morning and early afternoon with the researchers debating with the German delegation about crucial issues for the responsible use of AI. FGV's Director of Research and Innovation, Goret Paulo, describes the main points discussed at the meeting:

"The debate between FGV researchers and representatives of the German government on information integrity and the growing use of AI technologies made it clear that we need to invest in educating the population and regulating the use of these tools. In the academic environment, investment must focus on creating a culture and drawing up standards in order to promote the responsible and ethical use of artificial intelligence tools."

According to the Director, the dialog with the German delegation should continue soon, with a new meeting with representatives of the German government, with the aim of identifying possibilities for cooperation in scientific research.

Learn about other FGV projects that include applications of Artificial Intelligence tools:

Center for Analytics and Security Policy Studies Center for Analytics and Security Policy Studies (FGV Analytics)
Artificial Intelligence and Higher Education
LegalAnalytics

Visit the online portal of the schools and institutes that took part in the debate:

FGV Direito SP - São Paulo Law School
FGV EMAp - School of Applied Mathematics
FGV Direito Rio - Rio de Janeiro Law School
FGV EESP - São Paulo School of Economics
FGV EAESP - São Paulo School of Business Administration
FGV EPPG - School of Public Policy and Government
FGV ECMI - School of Communication, Media and Information
FGV IBRE - Brazilian Institute of Economics