According to their report, Megan Fox, the star of the recent Transformer’s movie, had her images reused most on the Web. This is despite the fact that she was 18th and 65th on the two lists.
However, the major story is not that the Web finds Megan Fox attractive, but that Attributor was able to locate some 2,547 copies of the estimated 220 images in the two lists.
Attributor appears to be advancing rapidly in the area of image detection and this could be a boon for visual artists, who frequently have struggled to locate their works on the Web.
Details and Serious Findings
The humorous nature of the study belies some very serious findings. In addition to the finding the more than 2,500 copies of the images, an average of over then copies per image, there were several other details to emerge.
- Some 73% of the sites that used the copied images were running ads, indicating that the use was commercial.
- A third of the sites were visited by more than 50,000 visitors in the month of December.
- Only 13% of the copies linked back to the original source.
- Copies for the top ten women outranked the original image in Google Image Search 100% of the time.
Though the idea that images of beautiful women will be widely copied on the Web should surprise no one, the nature of the use is very worrisome. By in large, the copies were both unattributed and commercial in nature, raising very serious concerns.
In short, these were not merely fans using the images, only 40% of the reuse was on either a fan site, blog or personal home page, but rather for-profit ventures exploiting the images for commercial gain.
In the end, only 7% of all sites that had copied images appeared to be licensing them appropriately.
Attributor’s image detection system is still being tested. Their site indicates that the service is not completely ready and we should not assume that this experiment is an indication of how well the final product will work.
However, if the results are accurate, it is clear that photographers and visual artists might want to start following Attributor more closely. Currently, there is no simple way to search for copies of an image on the Web. This is in stark contrast to writers, which have a slew of offerings to check for plagiarism.
Currently, nothing like this exists for visual artists and I anticipate that there will be a great deal of interest among photographers and artists alike when and if this system is made available to to the Internet at large.
Disclosure: I am a consultant for Attributor.