Conflicts are inevitable in collaborative work and can harm team outcomes when poorly managed. While AI is increasingly envisioned as a third party in teamwork, potentially acting as a customizable and knowledgeable mediator, its design and role in such situations remain unclear, especially regarding cross-cultural expectations. To explore these expectations, we conducted five speculative design workshops in Germany and Japan and synthesized scenarios, sketches, storyboards, and reflections. We use these materials as speculative probes into acceptable AI involvement in team conflict, identifying recurring intervention patterns around when AI should step in, which roles it may take, and what actions are appropriate. We also reveal cross-site differences: German participants more often envisioned low-salient AI guiding resolution, whereas Japanese participants imagined brief, bounded interventions by embodied AI. From these insights, we derive design lenses for AI-supported conflict mediation that preserve human agency and reflect on implications for cross-cultural Research through Design.
Chen et al. (Fri,) studied this question.