Speculating gender in conversational interfaces

The paper investigates gender biases and stereotypes in the design of conversational interfaces such as chatbots and virtual assistants. Recently the topic has received a growing attention in the design community. In our research, we focus on the issue of the lack of diversity and viewpoints that may unintentionally create services that are biased towards a particular group of users, while ignoring the needs and perspectives of others. To address this issue, we organised a speculative design workshop inviting 9 people with different gender and sexual orientation to reflect on gender biases in conversational interfaces with the ultimate objective of identifying recommendations for the design of more inclusive services. In the paper, we first review the state of the art on gender biases in conversational interfaces, later we introduce the theoretical framework that inspired the workshop design, namely feminist interaction design and speculative design. The description of the workshop follows, together with the presentation and discussion of results illustrated in the form of guidelines for conversational interface design that actively seek out and embrace diversity. These include concrete recommendations for adopting a more inclusive language to avoid stereotypes and considering the potential impact of the interface on different user groups. In the conclusions, we recommend engaging users in the design process who represent a wide range of demographics. By doing so, designers can gain a deeper understanding of the needs and preferences of diverse user groups and create more inclusive and equitable services that benefit everyone.