AI content labeling and user engagement on social media: The role of AI level, content type, and disclosure timing

Would you still be just as willing to like content on social media, such as Instagram, if you knew it had been generated by AI? This question is explored in a new article by Freya Seeger, Michael Wessel, and Christiane Lehrer in the journal Electronic Markets. In two online experiments involving nearly 700 participants, simulated Instagram profiles were shown whose content was labeled as human-created, AI-enhanced, or AI-generated. The result: the greater the share of AI involvement, the lower the users’ engagement. The effect is particularly strong for emotional content, while rational posts are less sensitive to it. The study therefore provides practical guidance for platforms and content creators on how to handle the increasingly regulated requirement to label AI-generated content.

Link: https://link.springer.com/article/10.1007/s12525-026-00883-2External link