Year: 2025
Type: dissertation
Abstract: Artificial Intelligence (AI) is reshaping organisations, creating opportunities alongside ethical, legal, and societal risks. While global guidelines exist, translating principles into daily practice remains a challenge. This thesis examines how organisational structures, such as AI ethics committees, AI Compliance and Risk Officers, and AI Champions, can embed responsible AI governance into routine operations. Building on prior work, the study combines theoretical and operational perspectives. A mixed-methods approach, including literature reviews, case studies, and document analysis, explores success factors and barriers in implementing AI governance. Findings show that committees and structures enable interdisciplinary oversight when integrated into decision-making, while designated officers connect governance decisions to operational execution. AI Champions promote awareness and communication within business units. The thesis proposes an integrated model aligning risk management, documentation, monitoring, transparency, and data governance with specific structures, adaptable across contexts. Beyond compliance, particularly with the EU AI Act, the model positions responsible AI as an ongoing organisational capability, fostering trust, reducing risk, and enabling innovation aligned with societal values.
Author: MALDONADO DIZ LATINI, LUCAS
Institution: University of Pisa
Language: Italian
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7112282941
