Social Explainability for trustworthy AI: What types of explanations can help users develop appropriate trust?

Bianca Slocombe, Frens Kroeger – Coventry University

The project proposed here brings a socio-technical approach to Explainable AI, in order to help users develop appropriate trust and avoid online harm. The proposed project will develop and test new types of explanations and trust building strategies that are comprehensible, relevant and “normal” to users in the context of their everyday life-worlds.