This workshop aims to trigger discussions on challenges and current trends on trusted AI. Firstly, a background of trusted and trustworthy AI will be given, with some of the key aspects of trusted AI systems being presented. Further, we address how to guarantee ethical characteristics of AI systems and how to design them according to system engineering principles. Next, we will exemplify some of the challenges we are currently facing in the context of generative AI and increasing complexity of AI systems, in a cross-domain perspective. Gaps which are currently not covered will be discussed. 

Objectives of the workshop

• Main objective of the workshop is to foster discussions and collaboration on mapping challenges and opportunities related to trust and safety in complex AI systems, with an emphasis on high-risk environments. The organizers are also involved in the AI Trustworthiness Characterization Task Group and AI Risk Activity Group, coordinated by DCU Adapt, TAILOR and Adra.

Agenda

• Short introduction, Rodica Mihai
• Principles of Trusted AI, Dr. André Meyer-Vitali
• Trust in the Age of Generative AI, Dr. Emmanuel Kahembwe
• Trust and safety in the context of complex AI systems: challenges related to verification and validation, Dr. Rodica Mihai
• Discussion session

List of the workshop speakers:

• Rodica Mihai, NORCE
• André Meyer-Vitali, DFKI
• Emmanuel Kahembwe, VDE