Background: AI-generated fiction has become increasingly readable, yet its literary acceptance depends on more than fluency, because readers must also perceive coherent organization, aesthetic purpose, and trustworthy narrative agency across platforms, classrooms, publishing experiments, and increasingly ordinary digital reading environments worldwide today. Objective: This study investigates how narrative coherence, aesthetic judgment, and reader trust interact in the reception of synthetic fiction. Method: This study applies a qualitative-dominant comparative design to twelve English-language AI-generated short stories, combining a six-dimensional coherence matrix, a six-dimensional aesthetic rubric, and a six-dimensional trust framework with preserved reader-ranking metadata. Results: Most narratives sustained event sequence, temporal continuity, entity stability, and closure, although causal linkage remained comparatively uneven. Aesthetic evaluation was weaker than coherence, particularly in emotional resonance, interpretive depth, and imagery, while originality and stylistic control varied across models. Reader trust was strongest at the level of narrative reliability but declined when authenticity, intentionality, literary legitimacy, continuation, and recommendation were considered. Implication: These findings indicate that readable synthetic fiction may remain only conditionally literary when formal coherence is not accompanied by aesthetic distinction and perceived purpose. Novelty: This study advances an integrated reader-centered framework that separates intelligibility, literary valuation, and trust while preserving their analytical interdependence.