Real versus AI-generated war videos: comparative effects on viewer attitudes, emotional arousal, and behavioral intentions

With the advent of generative AI technology, the massive production of synthetic imagery and videos of war that convincingly mimic the images of real wars is made possible through rapid dissemination through social media, posing critical concerns regarding the potential psychological implications of this form of media content. Studies on deepfake technology reveal that these kinds of media content can contribute to uncertainty, distrust in news, and even negative impacts on democracy, despite being less “persuasive” compared to other forms of media. Furthermore, continuous exposure to the disturbing image of war on the internet is linked to trauma, fear, and perception of danger. In the research study social media users were randomly assigned to watch either a real or an AI-generated war video pertaining to the Middle East crisis or the Russia–Ukraine war. Participants completed pre-exposure measures of media literacy, prior conflict attitudes, and psychological baseline, followed by postexposure assessments of perceived authenticity, trust, emotional arousal, agenda-congruent attitudes, behavioral intentions (sharing, policy support, donation willingness) and psychological distress.