Navigating the Paradox Between GenAI Driven Performance and Responsible AI Imperatives in Professional Services Firms

Details

Supervisors

Decaux, Loïc

Faculty

Louvain School of Management

Degree label

Master [120] en sciences de gestion, à finalité spécialisée

Abstract

enThis master's thesis investigates how professional services firms reconcile the adoption of Generative Artificial Intelligence as a performance driver with the imperatives of Responsible Artificial Intelligence. Faced with a rapid diffusion of these technologies that has largely outpaced the development of governance frameworks, consultants, auditors and advisors. They find themselves arbitrating in real time between productivity pressures and obligations of transparency, confidentiality and human oversight. The study examines four emerging organisational practices: architectural pre-emption through technical control, tiered model selection, friction by design, and ambassador networks bridging operational teams and central governance functions. These practices are analysed through academic literature and semi-directive interviews within eleven professionals working in Big Four consulting and audit firms, a financial institution, a telecommunications operator, an engineering firm, a railway infrastructure manager, and independent practice. The analysis highlights a logic of progressive layering, in which the most mature firms combine technical isolation, structured governance processes and cultural socialisation. The less advanced organisations still rely largely on individual judgement, which the literature consistently identifies as the most fragile foundation. The study finally outlines concrete directions to strengthen the balance between performance and responsibility within professional services firms operating in a technological and regulatory environment in constant evolution.