Artificial Intelligence on Social Media and User Responses

This research targets multimedia content generated by Artificial Intelligence (AI) on social media and measures user responses to the content. AI-generated content on social media has the potential to spread misinformation, so it is important to investigate the type of responses such content evokes. This research asks how easily users can recognize the provenance of AI-generated content, what emotional reactions they have to the content, and how factors such as disclaimers, topic, and platform effect recognition and reaction. The study was done by analyzing comments on popular posts on TikTok and X containing multiple types of AI-generated media spanning a wide range of topics. Findings underscore a dominant majority of negative responses (70.8%, 177 comments) and comments with themes of Aversion (45.2%, 113 comments). Contextual analysis pointed out a stronger negativity towards disinformative posts (89.2%, 33 comments) and more positivity towards humorous posts (39.3%, 11 comments). Differences between platforms showed that X users properly recognized the provenance of AI content 7.3% more than TikTok users, further influenced by the presence of disclaimers. User disclaimers were more effective than platform disclaimers, showing the pivotal role users play in combating misinformation on social media. This research displays the scarcity of platform-initiated disclaimers, showing a need for more proactive measures to identify AI content. 9.6% of responses (24 comments) included legislative sentiments, which paired with such a large majority of negative responses highlights public support for regulatory interventions as societal apprehension towards AI lingers. As AI continues to develop, more research is needed to determine the ability of humans to discern the provenance of AI-generated multimedia content, and new ways to combat misinformation on social media may be needed to address this new technology.