Facebook has asked its users in Australia to send in naked pictures and videos of themselves in a bid to tackle revenge pornography. Using the images, the firm is hoping to develop tech that could prevent ex’s uploading compromising media without permission in the future.
In short, the system takes the images that users send and creates a ‘hash’ fingerprint of them. To the outside observer, this code will appear as a string of letters and numbers on screen, but in practice will allow Facebook’s computer software to detect if the image has been uploaded to the site again. If the scheme is successful, further trials will take place in the UK, the US, and Canada.
The company is hoping to save users from the potential embarrassment of requesting Facebook to delete the images, and the agonising wait that this may entail.
The use of image-matching technology to fight online abuse is nothing new. In March 2012, Microsoft released the first image-matching software, PhotoDNA, to police forces in a bid to help them detect child pornography online. The tech is now used with Microsoft’s own online services – Bing and OneDrive – and has become a staple among other tech firms including Google Gmail, Twitter, and the National Centre for Missing & Exploited Children.
Julie Inman Grant, Australia’s e-Safety Commissioner, affirmed that Facebook will not store the images as after they have been processed into a hash – the code is all that will kept on-system. Grant added that the hash cannot be turned back into an image.
During the Australian trial run of the scheme, Facebook has noted that it will hold a blurred version of the images that users submit, to be reviewed by the customer support team, before hashing it. The firm claims that this is to ensure the tech is working correctly.
According to Sky News, there are fears that images sent in will be hacked and seized by cyber-criminals before they are deleted, intercepted on their way to Facebook, or that hackers find a workaround for the system, such as resizing the pictures. But Alex Stamos, Facebook’s Chief Security Officer, affirmed that the company are improving the technology to detect changes in the images. Meanwhile, a spokesperson for Facebook told the Telegraph: “This is an initial pilot in Australia. We look forward to getting feedback and learning.”