Explainable AI in practice: Build trust and encourage adoption

Applications rooted in artificial intelligence (AI) have stirred up a revolution in almost every industry. Not in the least, this includes the insurance sector, which has traditionally been highly data-driven. However, despite the great predictive powers these machine learning (ML) algorithms possess, a major hurdle limits their widespread adoption: they are often opaque black boxes. This lack of transparency limits both the public’s and your company employees’ trust in such algorithms. In this article we discuss the techniques available to mitigate this issue and what should be considered in their implementation.