Beyond human touch: evaluating the effectiveness of AI, human, and hybrid-generated tourism promotional texts

Purpose – This research aimed to compare the effectiveness of human-generated, AI-generated (ChatGPT), and hybrid promotional texts using the AIEDA model, which incorporates anticipated emotions from the Model of Goal-Directed Behavior. 
Design/methodology/approach – We conducted an experimental study and applied PLS-SEM and other inferential statistics to analyze whether 455 participants from the UK could differentiate between human-created, AI-generated (ChatGPT), and hybrid promotional texts and compared promotional effectiveness. 
Findings – Participants were not able to distinguish between the three texts. However, the human-generated text was consistently less preferred. Effectiveness perceptions declined as participants suspected AI authorship, regardless of the text’s actual source (human, AI, or hybrid). The AIEDA model, which incorporates positive and negative anticipated emotions, proved effective in all three scenarios. 
Practical implications – Tourism marketers can leverage AI for content creation but should maintain human oversight. 
Originality/value – The study adds to communication theories and research on promotional text by deepening our understanding of algorithmic authorship and its impact on consumer behavior in tourism marketing. This study was one of the first to compare the effectiveness of human-created, AI-generated, and hybrid texts in tourism marketing, correlate respondents’ preferences toward advertising texts with texts’ suspected AI authorship, and use an extended AIEDA model.