Seminar: Transparent AI in healthcare

The Health Law Research Centre welcomes you to sign up for a seminar with Inês Godinho, Associate Professor at Universidade Lusófona, Portugal, titled "Transparent AI in healthcare".

The core question for the seminar is what Transparent AI ‘should’ mean or stand for? Is it limited to the explainability criteria, or does it rather relate to oversight? Moreover: is there a difference between a transparency by design and a transparency on demand approach? In reasoning about these issues, it is necessary to have a multidisciplinary approach, crossing both ethics and law. 

Different ethics instruments frame transparency as a built-in requirement for AI systems or tools. Furthermore, the set of obligations and their types in the AI Act, although not actually providing meaning, are able to give insight on the need for a multifactor – or comprehensive – transparency approach. Inês Godinho argues that transparency, to come forth as an intrinsic quality, able to provide or foster trust in AI, cannot be confined to the link to explainability – even though it is a necessary link – but rather goes towards a meaning of awareness. She means that this awareness should not be mistaken for over-explainability – possibly leading to information overload – but that it can lead to more explainable AI (XAI) in certain areas, reenforcing a transparency by design approach in detriment of a merely transparency by demand approach.

You can sign up here!