Practical Application and Limitations of AI Certification Catalogues in the Light of the AI Act
Gregor Autischer, Kerstin Waxnegger, Dominik Kowald
TL;DR
The paper addresses the practical challenge of certifying AI systems under evolving regulatory regimes, focusing on applying the Fraunhofer AI Assessment Catalogue to an open-source emotion-recognition system embedded in a larger RIOT installation. It shows that while the catalogue provides a rigorous, multi-dimensional framework, its thorough, documentation-centric process is time-consuming and hampered by gaps when the target system lacks an active development team. The authors discuss the benefits and limitations of using open-source, real-world contexts for sample certifications and compare Fraunhofer with alternative catalogues (TÜV and Auditing ML Algorithms) to identify pathways for streamlining certification. The work offers actionable recommendations for future sample certifications and real-world practice, emphasizing complete system documentation, active development engagement, and the potential value of more structured, data-centric certification aids.
Abstract
In this work-in-progress, we investigate the certification of AI systems, focusing on the practical application and limitations of existing certification catalogues in the light of the AI Act by attempting to certify a publicly available AI system. We aim to evaluate how well current approaches work to effectively certify an AI system, and how publicly accessible AI systems, that might not be actively maintained or initially intended for certification, can be selected and used for a sample certification process. Our methodology involves leveraging the Fraunhofer AI Assessment Catalogue as a comprehensive tool to systematically assess an AI model's compliance with certification standards. We find that while the catalogue effectively structures the evaluation process, it can also be cumbersome and time-consuming to use. We observe the limitations of an AI system that has no active development team anymore and highlighted the importance of complete system documentation. Finally, we identify some limitations of the certification catalogues used and proposed ideas on how to streamline the certification process.