Paper Type
Complete
Paper Number
PACIS2026-1504
Description
Generative AI increasingly produces branded video content whose authorship is ambiguous, hybrid, or algorithmic, challenging how users interpret authenticity and agency in digital environments. This study examines how viewers construct and renegotiate authorship perceptions when encountering AI-enabled branded video artifacts. Drawing on 42 semi-structured interviews and an interpretive Gioia analysis, we identify three interrelated evaluative dimensions: cognitive and sensory experience, emotional and social engagement, and perceived authenticity and credibility. We further theorize attribution timing as a process variable shaping engagement trajectories. Two dominant interpretive pathways emerge: disclosure-triggered reframing, in which users retrospectively reinterpret content after learning it was AI-generated, and immediate AI recognition, in which attribution skepticism is embedded from the outset. By integrating attribution theory and affordance theory, this study reconceptualizes authenticity as a temporally negotiated sociotechnical judgment grounded in perceived agency rather than in artifact properties alone.
Recommended Citation
KEFI, Hajer; Oulevey, Frederic; and ABBAS, Noama, "Attribution Timing and Agency Construction in AI-Enabled Digital Artifacts" (2026). PACIS 2026 Proceedings. 8.
https://aisel.aisnet.org/pacis2026/ai_ml/ai_ml/8
Attribution Timing and Agency Construction in AI-Enabled Digital Artifacts
Generative AI increasingly produces branded video content whose authorship is ambiguous, hybrid, or algorithmic, challenging how users interpret authenticity and agency in digital environments. This study examines how viewers construct and renegotiate authorship perceptions when encountering AI-enabled branded video artifacts. Drawing on 42 semi-structured interviews and an interpretive Gioia analysis, we identify three interrelated evaluative dimensions: cognitive and sensory experience, emotional and social engagement, and perceived authenticity and credibility. We further theorize attribution timing as a process variable shaping engagement trajectories. Two dominant interpretive pathways emerge: disclosure-triggered reframing, in which users retrospectively reinterpret content after learning it was AI-generated, and immediate AI recognition, in which attribution skepticism is embedded from the outset. By integrating attribution theory and affordance theory, this study reconceptualizes authenticity as a temporally negotiated sociotechnical judgment grounded in perceived agency rather than in artifact properties alone.
Comments
01-AIML