Analysis Syntax and Cleaned Data

This study explores the use of AI within narrative messages about corporate social responsibility (CSR) initiatives, building on theories of narrative communication and expectancy violation theory (EVT). A 2 (message type: narrative vs. expository) x 4 (source label: human, AI-assisted human, AI, no source) between-subjects online experiment (<i>n = </i>277) investigated how various AI and human source labels used in narrative or expository messaging impacted stakeholder perceptions and behavioral intentions. Results indicate that AI source labels elicited lower perceived message transparency (accuracy and disclosure) and organizational transparency than human source labels but did not otherwise impact stakeholder perceptions or behavioral intentions. Moreover, exposure to narrative (vs. expository messaging) generated higher perceptions of message transparency (accuracy and clarity) and organizational transparency. Theoretically, these findings contribute to narrative and EVT research in strategic communication and provide a novel examination of AI disclosure in CSR storytelling. Practically, findings provide insight for practitioners regarding stakeholders’ responses, or lack thereof, to the disclosure of AI use in strategic storytelling.