What Exactly Is Being Supported? Target-Specific Responses to AI-Generated Political Videos on TikTok

Abstract Responses to AI-generated political videos can evaluate different objects. This qualitative study uses Content–Behaviour–Network (CBN) to examine six TikTok videos and 360 selected responses. Close readings distinguish appreciation of performance, informational commitments, political assent, and judgements of official communication. Comparing whole-response attribution with target-specific interpretation shows how collapsing distinct evaluative objects changes the political meaning attributed to a response. Platform labels and one critical re-presentation show how disclosure and critical framing add interpretive context as videos circulate. CBN connects these observations and ties each judgement to its target and source of evidence.