Trust in AI is about much more than model accuracy. It is affected by how well we understand it and how it reached its decisions, our control over inputs, its past record, and the consequences of it going wrong.
Trust can be undermined by many factors in data, design and use. Many companies have built good AIs, but users have issues trusting results.
Our whitepaper examines what makes AIs trustable, and explains how to create AIs that are: