Conversational user interfaces (CUIs) are rapidly changing how we engage with technology, impacting sectors like healthcare, education, finance, and entertainment through both embodied (e.g., social robots) and disembodied (e.g., chatbots) forms. Their effectiveness stems from advancements in language processing, voice recognition, and machine learning, enabling them to respond accurately.

As CUIs become more prevalent and powerful, it’s essential to address issues of bias and fairness. These systems can unintentionally learn and amplify stereotypes from their training data, leading to varied interactions across different user groups. This is particularly concerning in critical areas like healthcare and education, where biased responses can significantly affect well-being and access to resources.

Topics

This workshop aims to explore bias and fairness in conversational systems and invites submissions on related topics.

• Social and behavioural assessment of Bias & Fairness.

• Conversational design factors.

• Embodiment in CUI - the role of CUIs’ medium in shaping perceptions of Bias & Fairness (inc. social robotics).

• Mitigation strategies for Bias & Fairness.

• Metrics for measuring Bias & Fairness in CUIs.

• Behavioural paradigms for eliciting agents' behaviour or fixing agents' behaviour.

• Wide application context - CUI across different settings and environments (e.g., healthcare, education, etc).

• System and ML perspective mitigating bias and fairness concerns in CUI.

• Multimodal approaches to Bias & Fairness in CUIs.

• Ethical considerations of Bias & Fairness in CUI.

Papers

We welcome original work but also systematic and narrative reviews, theoretical propositions and hypotheses, as well as opinion pieces aiming to facilitate meaningful discussions across CUI researchers interested in Bias and Fairness. Papers can be either full papers following ACM CUI's guidlines, or short contributions of 2-3 pages (excluding references). Opinion and perspective papers can be of an exploratory or philosophical nature, as well as reflective writing on authors’ past works that relate to bias and fairness in CUI. ACM template link on Overleaf here.

 Accepted papers will be presented at the workshop, which will take place in person at CUI 2025, and will be published on the workshop's website. All submissions should use the default "sigconf" style. Submissions should be single blind, i.e. do not need to be anonymized.

Submission

Submit your paper by the deadline (check Important Dates) in 2nd June 2025.

Reviewing

All submissions will be reviewed by 2 reviewers.