ABSTRACT
Forecasting and assessment of societal risks related to digital innovation systems and services is an urgent problem, because these solutions usually contain artificial intelligence algorithms which learn using data from the environment and modify their behaviour much beyond human control. Digital innovation solutions are increasingly deployed in transport, business and administrative domains, and therefore, if abused by a malicious human activity or by a technical malfunction, may pose serious threats to public and to society. This chapter presents early results of testing a new approach to forecasting and prioritizing societal risks associated with deploying digital innovations. This approach – termed the RADI methodology – features two new elements: combining small expert panels with AHP-processed human assessments, and a relationship matrix visually mapping identified threats with available preventive policies, including the Socially Responsible Design approach. Although in an early version, the RADI methodology has been shown to be useful in guiding expert panels in profiling arrays of interventions to be undertaken by specific institutions against potential damage caused by non-validated and non-certified digital innovations.
