Paper accepted at AlpCHI'26
Enough With Trust!
Trust has become the default concept in human-AI research — the metric designers optimize for, the variable researchers measure, the goal everyone agrees on. But what if we've been asking the wrong question? This paper takes a critical look at how trust is used in the field, exposing methodological shortcomings, misinterpretations, and a narrow view of what makes an interaction "successful." It then proposes a shift toward frameworks that better capture what responsible and meaningful human-AI collaboration actually requires.
Congratulations Patricia Kahr and Abraham Bernstein for getting accepted! The paper Enough With Trust! Why We Must Move Beyond a Convenient but Insufficient Concept in AI-Supported Decision-Making Research will be presented in Ascona at AlpCHI'26, and can be found here.