Trust in Vision-Language Models: Insights from a Participatory User Workshop

TL;DR

This paper explores how users develop trust in Vision-Language Models through a participatory workshop, aiming to inform future trust assessment strategies and improve user interaction with these AI systems.

Contribution

It introduces preliminary insights from a user workshop to better understand trust dynamics in VLMs and guides future research on trust metrics and user engagement.

Findings

User trust in VLMs is complex and context-dependent.

Workshop insights highlight key factors influencing trust development.

Results inform future studies on trust measurement and user interaction strategies.

Abstract

With the growing deployment of Vision-Language Models (VLMs), pre-trained on large image-text and video-text datasets, it is critical to equip users with the tools to discern when to trust these systems. However, examining how user trust in VLMs builds and evolves remains an open problem. This problem is exacerbated by the increasing reliance on AI models as judges for experimental validation, to bypass the cost and implications of running participatory design studies directly with users. Following a user-centred approach, this paper presents preliminary results from a workshop with prospective VLM users. Insights from this pilot workshop inform future studies aimed at contextualising trust metrics and strategies for participants' engagement to fit the case of user-VLM interaction.