Artificial Intelligence (AI) has delivered unprecedented opportunities and also spawned new societal hazards in today’s rapidly evolving technological context. One of the most concerning threats involves the increasing prevalence of AI-generated political misinformation, which undermines democratic dialogue and public faith. This qualitative content analysis studied how this misinformation is created and shared in TikTok through five purposively selected videos, and audience comments on the videos. A two-cycle coding process, descriptive coding, and thematic analysis were used to conduct the qualitative content analysis. Inter-coder reliability was determined through Cohen’s Kappa (κ = .82), thus showing strong agreement, and reflexivity and anonymizing were used for rigor and ethics-related issues. The study revealed five interrelated themes that illustrate the layered strategies of misinformation: polarized narratives, which reduce politics to moral binaries of heroes and villains; engineered authenticity, which uses realistic visuals and synthetic voices to mimic credibility; linguistic persuasion, which employs rhetorical and emotional cues to foster skepticism; emotional amplification, which manipulates affective responses such as outrage or admiration; and credibility hybridization, which merges real and fabricated elements to blur truth and fiction. Collectively, these layers form a credibility stack, a multimodal structure combining narrative, visual, linguistic, and emotional manipulation to enhance believability. The findings underscore that AI-generated misinformation is more persuasive and difficult to detect than traditional propaganda, calling for strengthened media literacy initiatives and policy safeguards to mitigate its influence on democratic discourse.