Public acceptance has an important impact on the development of face recognition technology, but there is still a lack of research on the public acceptance of face recognition. A cognitive process model was constructed based on social exchange theory and responsible innovation evaluation criteria. The model examines the behavioral decision-making balance between perceived benefits and perceived risks to determine usage intention for face recognition technology. The structural equation model was used for empirical testing. The results indicate that perceived benefits and moral ethics significantly affect the public’s usage intention. However, in the face of the perceived risks, a phenomenon of “knowingly violate” rather than “timely stop” is observed. Previous privacy invasion experience, information sensitivity and negative news are found to increase perceived risks, though they do not outweigh the benefits of face recognition. Face recognition convenience is shown to enhance perceived benefits, yet it does not lead to the disregard of privacy leakage risks. Knowledge and security related to face recognition is positively and negatively correlated with perceived benefit and perceived risk, respectively. The findings contributes to a deeper understanding of the public concerns and needs for facial recognition technology, providing guidance for optimizing technology development and market promotion strategies.