Members of ELSA “European Lighthouse on Secure and Safe AI” met in Bowness-on-Windermere on the 25^th June to discuss the research being done in Work Package 3 of the project. This workshop was to examine the legal and ethical aspects of machine learning in the light of the EU Act.

Researchers from Work Package 3 partners as well as other experts in their fields from the European Commission, ELLIS and many different universities and industry came together to discuss how technical, legal and ethical considerations all need to be taken into account when trying to solve issues of machine learning with regards to the EU AI Act.

The event was organised by Lancaster University.

The day started with presentations from Lilian Edwards, Karen Yeung and Rotem Medzini. They discussed the legal aspects that needs to be considered when working in this field. They gave a comprehensive outline of the AI Act , talked about the background of global AI governance and discussed transparency and human oversight for high risk AI within the AI Act.

Dmitry Kangin then discussed the more technical aspects giving an overview of existing ML approaches.

We then heard from speakers who are working in this field and how they are tackling these issues with regards to Deep Fakes, Robotics and Document Intelligence.

The afternoon consisted of two brainstorming sessions where we discussed the obstacles that the current ML/AI solutions present in the light of the EU AI Act , whether this gap can be bridged and if so how?