Artificial intelligence is having a substantial impact. Breakthroughs in machine learning and advanced data analytics today can do everything from reducing carbon emissions to identifying potentially life-saving cancer treatments. Yet despite its popularity, many people -- including policymakers -- fundamentally misunderstand and mischaracterize the AI’s potential. Prominent and influential voices peddle negative visions of an AI future. This report addressed these debates by providing evidence and context within which such discussions are taking place. The ultimate goal was to provide a clear, unbiased view of the current state of AI as well as an outlook into the near future.

The Capstone team identified five applications of AI in detail: autonomous vehicles, autonomous weapons, financial risk pricing, customer recommendations and healthcare diagnostics. The paper examined the benefits and risks as represented by experts’ discourse and public opinion on the topic. It drew out mischaracterizations of AI and highlighted lessons learned from adverse outcomes. Finally, it provided an overview of relevant, current and pending AI regulation across the US, EU, Japan and China. The report included a specific set of recommendations for IBM and other industry leaders as well as policy makers, who, in navigating the AI landscape, need to ensure it is used responsibly, in a transparent and fully explicable manner, for the benefit of all people.