On January 30, 2020, the Centre for Information Policy Leadership (“CIPL”) at Hunton Andrews Kurth LLP submitted formal comments to the Department of Telecommunications at the Brazilian Ministry of Science, Technology, Innovations and Communications (“MCTIC”) on its public consultation on creating a national Artificial Intelligence (“AI”) strategy for Brazil (the “Consultation”).
CIPL welcomes MCTIC’s recognition of the dual objective of an effective AI strategy to not only address the risks and challenges presented by AI technologies but also to enable innovation and beneficial use of data through AI technology. CIPL believes that in order for Brazil to create a robust national AI strategy, it must prioritize privacy and data protection issues in the AI context.
CIPL’s comments focused on three of the eight specific thematic axes that were the subject of consultation—namely: (1) legislation, regulation and ethical use, (2) AI governance, and (3) international aspects.
In its comments, CIPL put forward several recommendations for MCTIC to consider with respect to a national AI strategy for Brazil, including:
- Acknowledging potential problems of algorithmic bias and discrimination in decisions made by autonomous systems and calling on organizations to continue working on solutions to address these concerns;
- Highlighting the importance of human intervention in AI contexts where the outcome of an automated decision implies a high risk of harm for the individual that is subject to it, while also acknowledging that organizations are best placed, and should have the flexibility, to define any methodologies and elements of review based on identified risks;
- Detailing various ways that explainability can be implemented in AI systems and encouraging organizations to develop further ways to facilitate transparency as it relates to AI outcomes;
- Calling for a layered approach to AI regulation should Brazil ultimately choose to regulate AI technology—this would involve looking at the rules contained within the Brazilian LGPD and its intersection with AI technology, leveraging and incentivizing accountable AI practices of organizations, and fostering innovative approaches to regulatory oversight;
- Emphasizing the importance of organizational accountability in achieving AI governance;
- Calling for the creation of an AI regulatory sandbox by the Brazilian Data Protection Authority (the “ANPD”); and
- Calling for collaboration with many different international partners in developing the responsible use and deployment of AI, including through international fora (e.g., the Organization for Economic Co-operation and Development), and regional groups (e.g., the European Commission High Level Expert Group on AI or the Council of Europe Ad Hoc Committee on AI), research institutions, universities and multinational corporations.