Semester of Graduation
Spring 2026
Degree
Master of Mass Communication (MMC)
Department
Manship School of Mass Communication
Document Type
Thesis
Abstract
This study investigates how the attribution of a message source (Human, AI, or Hybrid) influences consumer perceptions of authenticity and brand trust within two distinct communication contexts: public relations and advertising, in a condition of non-disclosure reflective of the current regulatory environment. The study utilizes a 3x2 between-subjects factorial design (N = 418). Participants were exposed to communication materials from a fictitious brand and evaluated them for perceived authenticity, brand trust, and AI suspicion.
Key findings indicate that human-generated content remains the most authentic to audiences, earning significantly higher perceived authenticity scores than synthetic communication. Results also identified a paradox in co-created content, in which content created by AI and refined by a human strategic communication practitioner resulted in the lowest levels of perceived authenticity.
Additionally, a full mediation analysis revealed that the trust gap between human and AI sources can be primarily attributed to the level of perceived authenticity from the communication material, not from a reaction to the technology itself. AI suspicion also served as a strong negative predictor, explaining approximately 24% of the variance in brand trust. These results extend the application of the Computers as Social Actors (CASA) and Human-Centered AI (HCAI) frameworks, suggesting that practitioners should prioritize human sincerity and verisimilitude over the efficiency gains brought by co-creation to maintain brand reputation in the current, opaque, media environment.
Date
3-27-2026
Recommended Citation
Lofton, Jacob A., "Selling Synthetic: Trust and Authenticity in AI vs. Human Strategic Communication" (2026). LSU Master's Theses. 6382.
https://repository.lsu.edu/gradschool_theses/6382
Committee Chair
Nihar Sreepada
LSU Acknowledgement
1
LSU Accessibility Acknowledgment
1