Artificial Intelligence (AI) - the Need for New Safety Standards and Methodologies

There has been a series of challenges to develop appropriate safety standards and methodologies as technology evolves, to ensure their safe implementation. These challenges, which first arose at the dawn of the industrial revolution, will inevitably continue. New technologies will always seek to forge ahead in a competitive marketplace, failure to do so will inevitably lead to organisational demise. However, such developments must be matched by a complement of research activity seeking to ensure that appropriate new safety standards and methodologies are put in place to maintain acceptable levels of risk. A new challenge now confronts us in the form of artificial intelligence (AI) where we stand at the frontiers of decison making in relation to what roles machines and humans should play with regard to optimal decision making and how these impact on safety. These impact on both, organisations which propose AI application and those which are responsible for setting standards and regulation. This is the subject for discussion in this paper.