Evaluating the Risks: Assessing the Impact of the EU AI Act on Artificial Intelligence Systems at High Risk

Abstract

This thesis examines the risks associated with high-risk artificial intelligence (AI) systems and assesses the effectiveness of the European Union's AI Act in mitigating these risks. Through a combination of theoretical literature review and empirical research, the study develops a framework that categorizes AI risks and mitigation strategies. Initially, the framework was informed by literature identifying risks like privacy violations, decision-making biases, and
compliance challenges, along with corresponding mitigation strategies such as data protection and transparent algorithms. Empirical research involving interviews with professionals across healthcare, logistics, and financial technology sectors revealed additional risks and practical mitigation strategies, which were integrated into the final framework. This research highlights the complexities of applying the EU AI Act across various industries,
uncovering a significant gap between theoretical intentions and practical implementations.
The findings suggest that while the EU AI Act provides a foundational framework, its
effectiveness is contingent upon adaptations that address the unique operational needs of
different sectors. The thesis underscores the necessity for dynamic regulatory approaches
that evolve with technological advancements and sector-specific requirements.

Date of Award	24 Jun 2024
Original language	English
Supervisor	Samaneh Bagheri (Examiner) & Vanessa Dirksen (Co-assessor)