Bots with feelings: Study explores how human customers react to AI chatbots with emotions

Artificial intelligence chatbots that show positive feelings - such as adding an -I am excited to do so!- or a few exclamation marks - do not necessarily translate into positive reactions or contribute to higher customer satisfaction, according to a recent study by researchers from the University of South Florida, the Georgia Institute of Technology and McGill University. As online retailers increasingly use artificial intelligence chatbots to streamline customer service tasks and replace their human counterparts, the researchers examined how emotion-expressing AI chatbots can impact customer service. -It is commonly believed and repeatedly shown that human employees can express positive emotion to improve customers- service evaluations. However, our findings suggest that this conventional wisdom does not necessarily apply to the case of an AI service agent,- said study co-author and McGill Professor Elizabeth Han , Desautels Faculty of Management. The research is particularly relevant with the rise of emotional AI - the branch that deals with processing and replicating human emotions - and the recent claim from a Google engineer that an unreleased AI chatbot was -sentient. That claim and the resulting backlash led to the employee's firing for violating employment and data security policies. In addition, a national debate ensued around what -sentient- means and whether Google's chatbot has consciousness or feelings.