AI as a creator or a helper: how disclosure and source credibility shape public perception of political advertising messaging

The rapid integration of AI into political communication has raised concerns about transparency, credibility, and accountability in election campaigns. Yet little is known about how disclosure of AI involvement shapes audience responses, particularly when perceptions are influenced by source credibility. This study addresses this gap through two experiments. Study 1 shows that when source credibility is low, ads disclosed as AI-generated are evaluated less favourably than those without disclosure. Study 2 extends these findings by comparing levels of AI involvement, revealing that AI-assisted ads generate more positive attitudes than fully AI-created ads, especially under low-credibility conditions. Results further demonstrate that perceived intrusiveness mediates these effects, with AI as the sole creator heightening resistance while AI as an assistant reduces it. These findings challenge the assumption that transparency uniformly enhances trust, highlighting instead the importance of framing AI’s role. The study advances theoretical understanding of AI’s impact on political persuasion and offers practical guidance for designing disclosure strategies in political campaigns.