I would like to ask if it's possible to prove that your image went through a DLNN like the photos did here with Ever. Say Ever never kept any photo access logs and decided to delete the model training logs or didn't keep logs for that at all. If all you know is that your photo was on their service, and you had their deep learning model (eg. as evidence in court), could you prove that your image was used to train the network?
Google has been doing this with Google Photos since its inception, the only difference being that Google actually uses it (in my opinion) for good by auto-tagging and offering search across your entire life of taking photos.
