Image tags are very important for indexing, sharing, searching, and surfacing images with private content that needs protection. As the tags are at the sole discretion of users, they tend to be noisy and incomplete. In this paper, we present a privacy-aware approach to automatic image tagging, which aims at improving the quality of user annotations, while also preserving the images' original privacy sharing patterns. Precisely, we recommend potential tags for each target image by mining privacy-aware tags from the most similar images of the target image obtained from a large collection. Experimental results show that privacy-aware approach is able to predict accurate tags that can improve the performance of a downstream application on image privacy prediction. Crowd-sourcing predicted tags exhibit the quality of the recommended tags.