The credibility cocktail on Instagram: Assessing the impact of source, verification label and image manipulation on news message credibility

In the current disinformation era, where both people and generative AI systems can easily create content, concerns about the credibility of online content have come to the fore in public and policy debates. Audiences are faced with the challenge of determining which news messages are credible and which are not. It remains however unclear how audiences evaluate the credibility of social media posts. This study evaluates three key ingredients of the credibility cocktail on Instagram: the source, the presence of a verification label and the use of images. Using two between-subject designs ( N = 963), participants were exposed to an Instagram message regarding climate change. The findings demonstrate that audiences primarily assess the credibility of messages based on the source, i.e. the person or organization sharing the information. Messages sent by both a news outlet and non-governmental organization are considered to be more credible than messages shared by dubious sources. Remarkably, a verification label next to this source has no impact. Hence, audiences do not trust so-called ‘blue check accounts’ more than accounts without a verification label. Moreover, AI-generated images are perceived as more credible than photoshopped ones.