Visual Cues in AI-Mediated Communication: How Anchor Attire and Gender Shape Expectancy Congruence, Trust, and Continued Viewing

Synthetic news anchors are increasingly deployed, yet little is known about how their designed visual attributes shape audience response. Because every attribute of an AI anchor is designed, a single visual cue can be isolated. This study asked whether attire matching the broadcast context raises perceived congruence, and whether congruence carries downstream effects on continued viewing. In a mixed-design experiment, 397 U.S. adults were assigned to one of eight conditions crossing anchor gender with the attire worn in each context, and viewed both an awards ceremony and a Christmas festival report rendered with photorealistic MetaHuman stimuli; each context was analyzed as a 2 × 2 design. Attire had no statistically significant total effect on viewers’ willingness to keep watching. What attire did change was appraisal. Contrary to a context-matching account, business attire produced higher perceived congruence than thematic attire in both contexts—the suit-schema paradox, so named because the attire designed to fit the occasion was judged to fit the broadcast less well. Neither context isolates fit to the occasion: one thematic garment was a costume, the other less reliably recognized. In an exploratory test the effect was present for the male anchor and not detected for the female anchor; that this reflects anchor gender is a hypothesis the stimuli cannot adjudicate, as each gender had a single face, voice, and business garment. Among the measured variables, congruence predicted AI trust, which predicted continued viewing intention; a parallel route through immersion is reported as exploratory on reliability grounds.