On November 13, experts discussed criteria for trustworthy AI systems at the acatech Forum Munich. The focus was on applications ranging from medical diagnostics to corporate management. In front of 40 participants, IBM expert Andrea Martin emphasized the importance of human control in AI decisions.
Carolin Anderson, project manager at MISSION KI, opened the event with an introduction to standards for fair AI. Moderated by Dr. Katharina Kaufmann, Andrea Martin (CTO Ecosystem & Associations DACH, IBM), Prof. Dr. Gitta Kutyniok (Chair of Mathematical Foundations of AI, LMU Munich) and Tim Vallée (Co-Founder of Beyond AI Collective) discussed key issues for the future.
Medical AI under human control
"In medicine, the final human review is indispensable. It ensures that decisions are not only technically correct, but also ethically justifiable", emphasized Martin. Diagnostic tools that support doctors but do not make independent decisions are seen as a role model.
Focus on fairness and transparency
The discussion focused on fair AI systems without discrimination. The experts pointed out that AI not only harbors potential biases, but can also help to make human prejudices visible and reduce them.
The following emerged as key success factors: the close integration of human expertise and AI technology, traceable data sources and decision-making processes as well as systematic training in the ethical use of AI.
Practice-oriented solution approaches
Another focus was on the formalization of trust standards. Automated verification procedures were presented as a promising approach, especially for small and medium-sized companies, to comply with legal requirements in a cost-effective manner.
Want to know more?
Interested in how AI is shaping our future? Take part in our upcoming events and actively shape the debate! You can find more information on upcoming events under News on our website. If you have any questions or suggestions, please contact us at kontakt@mission-ki.de.
