Facebook Introduces A New AI To Deal With Revenge Porn

Facebook is dealing with revenge porn by employing a new technology which can identify and remove sexually explicit pictures published without the permissions of the person or people appear in them.

Facebook’s current system relies heavily on the reports of people falling victim of revenge porn. After moderators are alerted of the cases, they will start reviewing them. It even puts forward a proposal that users supply Facebook with their own sensitive photos so that it can prevent revenge porn beforehand. However, as the largest social media site is not exactly a shining example of privacy, its suggestion faces fierce opposition from users.

With this new technology, artificial intelligence is used to identify suspicious pictures and then send them to human moderators for review.

Facebook has been struggling to deal with criticism regarding its failure to remove posts that go against its regulations or even eliminating valuable photos of art or history. With the pressure from the public, Facebook is working on enhancing its own performance in keeping its integrity.

This technology will be applied to both Facebook and Instagram. It has been under training to recognize revenge porn base on confirmed photos of this act. With this kind of training, it will flag an image of people in their undergarment for example, with text delivering insulting messages, as a potential act of revenge. In the US alone, there are more than 42 states have outlawed revenge porn. New York even allows victims to sue the culprits.

Facebook’s COO Sheryl Sandberg said that this technology was to solve the issue beforehand. She shared that users might submit photos they think might be spread online. Facebook will check if the image is online or not and then delete it permanently.

Facebook does not hope that this technology can be 100% effective, so it calls for users’ collaboration to report if they see anything that can be an act of revenge porn.