Regulation-by-Design Strategies in the Governance of Safe AI in Highly Regulated Sectors in Brazil
Artificial Intelligence (AI); AI Regulation; Regulation-by-Design; AI Governance; Bioinformatics; Compliance as Code
The proposed research is intrinsically interdisciplinary between the areas of Law and Computer Science. The candidate's academic path is marked by the intention to move beyond the analytical role common in the legal field, towards an acting role as a governance architect capable of directly interacting with software engineering and data science. The work addresses the gap in the industry and literature regarding the translation of abstract ethical-legal principles into technical and auditable requirements, focusing on sectors with specific regulations, such as biotechnology. Consequently, the proposed thesis significantly broadens the initial scope , centered on the development and application of the "Regulation-by-Design" methodology. This approach advocates that regulatory and compliance analysis should not be a final step, but a continuous process, starting from the conception of the technological project. The thesis investigates the translation of laws into software requirements ("Compliance as Code") using a risk-based taxonomy. The central objective is the transversal validation of the developed methodology by applying it to highly regulated areas in the country, such as health (applied bioinformatics), innovation, and the financial sector, which have high potential for the application and use of AI technologies