Transparency is frequently cited as a key ethical requirement for artificial intelligence (AI), yet its meaning remains ambiguous and often defined through a top-down technical or regulatory perspectives. The XpliCAI project adopts a user-centered participatory design approach to explore how healthcare stakeholders perceive and interpret transparency of AI systems. This paper reports findings from the first of six planned workshops, combining a preliminary questionnaire (N = 29) with four participatory focus groups involving healthcare professionals, students, patient partners, and caregivers.