Deepfake videos are highly realistic manipulated videos, blurring the line between what is real and fake. Deepfakes are generated with neural networks. The technology rapidly evolves, making...Show moreDeepfake videos are highly realistic manipulated videos, blurring the line between what is real and fake. Deepfakes are generated with neural networks. The technology rapidly evolves, making deepfake videos increasingly realistic. This makes it more difficult for humans to distinguish deepfake videos from real videos. Studies show that a large proportion of people are not familiar with deepfakes. Next, studies show that familiarity with a person portrayed on a deepfake video plays a role in detection performance. One of the threats deepfake technology brings is generating realistic fake news. The technology can be used go generate highly realistic fake news, which brings significant threats to society. This raises questions about the human recognition of deepfake videos. Most research on deepfake videos is focused on the algorithmic detection of deepfakes. Less is known about the human recognition of deepfake videos. The aim of the current study is to investigate the human performance at recognizing deepfake videos, and its possible predictors. It is hypothesized that social media use, conspiracy mentality, age and familiarity with a person portrayed on a deepfake are correlated to the human performance at recognizing deepfake videos. Our findings suggest that humans perform better at recognizing deepfake videos of familiar persons compared to deepfakes of unfamiliar persons. Next, our findings suggest a positive relationship between time spent on social media and performance at recognizing deepfake videos. No significant correlations were found between age, conspiracy mentality and deepfake detection performance.Show less
