The photos you upload online are not really secure, and not because someone can take them and use them, but because the NSA probably collects everything that is available.
This probably comes as no surprise, especially after a year of shocking revelations. What is surprising, however, is that the global surveillance system the organization has put in place not only collects photos but also knows who they belong to, thanks to its sophisticated facial recognition program.
According to The New York Times, the spy agency relies on facial recognition technology to exploit the photos it collects as well as emails, text messages, social media conversations, video conferences and other online communications.
Documents leaked by Edward Snowden show that agency officials are convinced that technical advances in facial recognition could revolutionize the way the NSA identifies targets around the world.
A file dating back to 2011 shows that the agency collects millions of images every day, including 55,000 photos that are perfect for facial recognition software, which represents a “huge untapped power.”.
Considering that in recent years the “selfie” phenomenon has flourished, it’s safe to say that the NSA is collecting many more images that are ideal for its programs.
The software developed by the NSA is quite advanced. According to the report, a PowerPoint presentation from 2011 shows photos of a man with and without a beard in different settings and about twenty-four data points about him, including his passport, known associates or suspected terrorist ties, and comments from informants.
Unfortunately for all involved, the US has no federal privacy laws to offer protection for photos, nor do surveillance laws within the country cover this issue, much less in the rest of the world.
The agency has stated, however, that the NSA does not have access to photos found in databases such as driver's license or passport photos of Americans, although there is no mention of whether or not the agency has access to the State Department's database containing photos of visa applicants.
If you're wondering how effective the NSA's system might be, it's worth mentioning that Facebook has a similar tool, although it's primarily used for tagging images. DeepFace, the company's 3D facial feature mapping software, converts them into a flat model before filtering the results based on a series of signals.
Facebook has managed to perfect the tool to provide results with 97.25% accuracy. Considering how long the NSA has been working on the same tool, we can say that theirs should be just as effective.
