Building a responsible AI ecosystem: insights from Brazilian initiatives

The growing use of Artificial Intelligence (AI) by companies in their pursuit to provide next-level products and services has spurred governments to discuss regulation of this technology with the aim to safeguard individual rights and national values, while fostering innovation and research.

Given the urgency of the matter, a collaborative effort has emerged both nationally and internationally to establish a governance ecosystem to ensure harmony between AI development and its responsible use.

In this context, noteworthy initiatives have emerged on a global scale. The World Economic Forum created the AI Governance Alliance, comprising a diverse group of experts and representatives from multiple countries, with the objective of promoting discussions on the safe, inclusive, and transparent development of AI.

The Association of Southeast Asian Nations (ASEAN) developed the Guide on AI Governance and Ethics to assist member countries in formulating their national AI development and regulatory strategies. In addition, the OECD created the AI Policy Observatory, which conducts studies and provides guidance on AI regulation, serving as a reference for many countries, including Brazil.

In the Brazilian context, two initiatives are worthy of mention. The first is Bill of Law 2338/23, stemming from the Senate’s effort to consolidate all legislative proposals and encourage a broader debate on AI regulation. The second is the Brazilian Artificial Intelligence Strategy (EBIA), drawn up in 2021 by the former Ministry of Science, Technology and Innovation (MCTI)

Among the various issues covered by the bill, the most notable is the indication of responsible governance of AI systems as an objective and principle of the law. It also mandates governance structures for AI systems, requiring AI developers to establish management and accountability frameworks and minimum content requirements.

Another policy related to AI governance in Brazil is the EBIA, developed as part of the Brazilian Strategy for Digital Transformation (known as “E-Digital”), established by Decree No. 9,319/2018.

Among its nine pillars, AI governance is highlighted as one of the main strategies for the advancement of the technology in the country. In the document outlining the plan, the MCTI detailed several strategic actions for AI governance, including: (i) improving access to open government data, (ii) advocating for data and model sharing between organizations in compliance with the General Law on the Protection of Personal Data – “LGPD”, and (iii) encouraging the development of risk management standards. The strategy also proposes establishing ethics and review boards for AI data within organizations, as well as implementing accountability practices. The EBIA includes the development of guidelines for data protection impact reports, in accordance with the provisions of the LGPD. Furthermore, the EBIA envisions educational initiatives and multisectoral dialogue to standardize governance practices.

Although the policy was created in 2021, the allocation of funds by the Brazilian government for this purpose is still insignificant. The Federal Court of Accounts (TCU) identified EBIA’s lack of initiative concerning the implementation of this policy in a report made public in May 2022. Consequently, the MCTI is in the process of reformulating the strategy, which shall be published soon.

Despite the challenges, ongoing efforts and future plans suggest a positive trajectory related to the implementation of AI governance in Brazil. The development of a regulatory and operational framework that supports both technological innovation and ethical and social responsibility are crucial for the country to secure its position as a leader in the responsible development and use of AI in the digital age.