Facebook rolls out new tools to fight "revenge porn"

Facebook will apply artificial intelligence technology to try to root out so-called “revenge porn” posted on its site before that content spreads widely, CEO Mark Zuckerberg announced Wednesday. 

“Today we’re rolling out new tools to prevent ‘revenge porn’ from being shared on Facebook, Messenger and Instagram,” Zuckerberg wrote on Facebook. “Revenge porn is any intimate photo shared without permission. It’s wrong, it’s hurtful, and if you report it to us, we will now use AI and image recognition to prevent it from being shared across all of our platforms.”

Revenge porn is an increasingly common form of sexual exploitation that takes advantage of the ability to share photos and videos online. With one click, a former partner — perhaps motivated by spite or seeking to blackmail an ex — can share intimate photos not only with his or her personal network but with far-reaching members of the public on these platforms. 

Facebook already applies image recognition to many other functions on its platform. For instance, it auto-recognizes familiar faces in new pictures and prompts users to tag those friends. 

Facebook has an existing policy to remove content that’s “shared in revenge or without permission from the people in the images.” 

As of now, users can flag content on Facebook, checking boxes like “I think [this] shouldn’t be on Facebook” and “It’s sexually explicit” before submitting the content for review. 

In recent years, Facebook has sought counsel from outside groups on how to fight back against the violence, harassment, and sexual exploitation that can dominate social media; for instance, Facebook has worked with the Anti-Defamation League on how to identify hate speech and with the National Network to End Domestic Violence on how to improve safety settings for survivors of abuse.