Harassment and offenses to conversational agents

: the case of BIA, the feminist chatbot

Abstract

Purpose: this investigation aims to understand if chatbots can effectively induce prosocial attitudes and behaviors in users, throughout the customer journey in order to effectively help to educate the society as a whole. Methodology/approach: this case study is about the specific case of bia, a chatbot from bradesco, with almost 1 million clients. A sample of 10 brazilian people participated in this study. Thus, quantitative data was collected through interviews. Findings: the results showed that, being used in many different areas, chatbots can actually be a source of inspiration and public opinion making, working as a digital influencer. In fact, it is possible to induce pro-social attitudes and behaviors in users, mainly on the prepurchase and post-purchase phases of the customer journey. Thus, it was possible to conclude that it is possible to educate the population through a robot and that the population, in general, is available to learn in this way. Originality/value: we believe that this study adds a lot of value and has a huge amount of originality since, in fact, and to my knowledge, we are the first one deeply studying the social dimension of a chatbot, which is not, from the outset, expected. In fact, and by studying the case of bia, this study was able to conclude that, as the process unfolds, bradesco gained notion that bia could effectively be a digital influencer, with the power to educate the (brazilian) population on issues, and in a first phase, essentially related to harassment, bullying and which always end up intertwining with the empowerment of women.

Date of Award	12 Oct 2022
Original language	English
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Susana Silva (Supervisor) & Carla Martins (Co-Supervisor)