Chia Sheng Chen

Type of the article: Research Article

Foreign information manipulation and interference (FIMI) increasingly operates in an information environment shaped by artificial intelligence, cross-platform circulation, and long-term narrative exposure. This paper develops an audience-centered relational framework for conceptualizing how AI-enabled FIMI may impair institutional trust in democratic societies. Rather than treating FIMI primarily as a problem of false content or message diffusion, the study examines how manipulative narratives may work through five interconnected dimensions: audience segmentation and targeting, narrative positioning, involvement and persuasion pathways, touchpoint configuration and information journey, and relationship degradation and trust impairment. Methodologically, this is a conceptual and theoretical study. It uses publicly available official reports, policy documents, public analyses, media reports, and think-tank publications concerning defense-related information narratives in Taiwan from 2022 to the first half of 2026 as illustrative materials, rather than as a systematically collected empirical dataset. The Taiwan case is therefore used as a theory-guided illustrative application, not as an empirical test of causal effects. The framework illustrates three possible trust-impairment pathways: weakening confidence in government competence, eroding credibility in democratic institutions, and reducing public identification with collective resilience. The study contributes to FIMI research by shifting attention from isolated messages to cumulative relational processes through which AI-enabled narrative environments may contribute to the gradual weakening of democratic trust.