Successful AI based apps by Facebook to battle revenge porn

With widespread of the internet, sharing of intimate videos and sexual images without the consent – commonly termed as revenge porn has been severely haunting millions of users worldwide. Facebook – the leading social networking site that has been constantly blasted for its loopholes in privacy and data protection practices have launched an AI based to detect revenge porn.
 
Facebook Inc. recently announced their strategy of launching a powerful technology tool that will implement both artificial intelligence and machine learning techniques to record and block revenge porn shared without the people’s consent across Facebook and Instagram.

If the image or video is being found breaching the Facebook’s Community standards, it will be discarded and in most cases, the social media site will also deactivate the account.  Besides helping users in reporting revenge porn, Facebook’s new advanced and sophisticated AI driven applications in tandem with the pilot system aim to build a proactive reporting tool to detect intimate images and block them before they spread, so it never gets shared across its platforms. Unlike current filters, this artificial intelligence tool effectively helps to detect near-nude content.

This Menlo Park, California-based company founded by Mark Zuckerberg, first started a pilot system last year to detect and flag revenge porn, but the system demanded a scary move from the user – the person has to first upload his own intimate images so that the system could study what kind of images needs to be blocked. After the user uploads his images of the pilot system – Facebook creates a digital fingerprint of the image – an encrypted version of the raw data in the image file. Therefore, when someone tries to upload these images they will be blocked from ever being shared on the platform.

However, the team handling the pilot system soon realized that the victims are often afraid and hesitant to report the revenge porn by themselves. In most of the cases, the victim is totally unaware that such content has been shared on the social network. Though a controversial move, the program gained a lot of positive support and response from victims and support groups. The new AI powered tool to detect and block non-consensual images, however, doesn’t require the victim’s involvement, unlike the pilot system.



In addition to these applications, Facebook is also set to launch a support platform called "Not without My Consent" on its safety page for users (victims) whose intimate images have been uploaded and shared without their permission. This support hub is primarily an advice information pool with lots of resources on what to do and whom to contact if the user is a victim of revenge porn. Facebook has also announced that in the next few months, it is set to update the support platform with various organizations and helplines such as SaferNet (Brazil), Professor Lee Ji-Yeon (South Korea), Revenge Porn Helpline (U.K.), Digital Rights Foundation (Pakistan) to offer psychosocial support.

More often, revenge porn has affected people especially women with different devastating consequences like anxiety, depression and other traumatic stress disorders leading to mental and health imbalance. On the professional work front, people are likely to lose their jobs and have a damaged relationship with their fellow colleagues. Considering revenge porn as a critical matter, the social networking giant decided to implement an AI powered tool to address such issues and to help curb the spread of revenge porn.

In the past five years, nearly 4 states have passed laws criminalizing revenge porn. Seeing the growth of revenge porn crimes, New York's law, which passed in February also permits the victim to file a lawsuit against the offenders and makes the crime a non-indictable offense.

Facebook the world’s largest social media networking platform with over 2 billion users has been roundly criticized for the last couple of years for allowing offensive posts to stay online for too long, massive data breach, exploiting user’s personal information and allowing users to post fake news. As per a detailed report by Edison Research – for the first time in the last 10 years, Facebook witnessed a drop of 5% in usage in the US. The numbers of users dropped from 67% to 65% indicating that the users are losing trust in Facebook.

Moving forward, Facebook CEO Mark Zuckerberg plans to implement a research-based approach that concentrates on the victims and focus on the following three key principles:

  • Build a clear support hub to help and encourage victims in reporting a violation
  • Develop proactive tools to detect and block someone from sharing non-consensual images
  • Give the complete power and freedom of online space to the victims so that they feel safe

The CEO did not mention when these new initiatives will be implemented but the software outsourcing company is striving extremely hard to have a more privacy focused platform. Going forward, Facebook also plans to merge all its messaging apps so that the user can easily connect and communities with their friends.