To trust or not to trust: When should AI be allowed to make decisions?

07.-09.09.2022: Workshop on ethics of AI in healthcare

The integration of Artificial Intelligence (AI) into clinical care evolves and progresses rapidly, moving from diagnostic image analysis in radiology and dermatology to ever more complex applications such as predictions in intensive care units or psychiatric care. The most intricate ethical challenges arising from such AI systems in medicine are linked to epistemological questions, especially when the very design of a system renders it opaque to human understanding and threatens trust in its use.

While opaque AI systems such as deep learning models pose problems for many of its potential applications in society, their impact on medicine raises specific questions that require careful philosophical, ethical, and legal deliberation:

  • How can informed consent, the bedrock of medical ethics, be obtained if a system is in principle unintelligible to its users?
  • How can accountability be attributed in the complex field of interactions between patients, medical professionals, formal and informal caregivers, AI developers, and the AI system itself?
  • What is a useful conceptual framework for trust in an AI system in medicine?
  • And ultimately what makes trustworthy AI in medicine?


With the goal of enhancing their scientific collaborations on the ethics of AI systems, the College of Humanities  at EPFL, the Digital Society Initiative at the University of Zurich, and the Swiss AI Lab IDSIA USI-SUPSI are organising in Lausanne on September 7-9 2022 an international meeting aimed at  bringing together philosophers, computer scientists, engineers and medical professionals to discuss and outline possible answers to the aforementioned crucial questions.

 

Submission of abstracts:
In addition to the invited speakers, we invite the submission of abstracts for the meeting from early career scholars (students, postdoctoral researchers, and junior faculty). Abstracts should be suitable for a 30-minute talk and are to be submitted by July 7, 2022.

 

Further information on the event and program details: here.