Earlier this week, four teenagers tortured and taunted an 18-year-old mentally disabled man and streamed the entire episode on Facebook Live. The suspects, all black, have been charged with a hate crime both for shouting epithets about white people in the video and because the victim is both white and mentally handicapped.

The horrifying incident has been roundly condemned by individuals from across the political spectrum, even the President of the United States. However, much less discussed has been Facebook’s role in accidentally promoting this kind of violent content — and the fact that the site allowed the torture video to remain posted for 30 minutes.

The social media giant did eventually take down the graphic video, noting that it violated the site’s rule against celebrating or glorifying crime. However, violent videos are allowed if the user who reposts it is condemning the violent actions. It’s also unclear what role the video played in the suspects’ eventual capture, though the video could be a powerful tool for providing the data of criminals to law enforcement.

“Nobody wants individuals to use their products for senseless acts, such as this,” Chris Michaels, director of communications at Wowza Media Systems, told Mashable. “Fortunately, the use of AI can help us begin to filter some of the hideous sides of humanity, and provide a safer solution for free discourse without egregious nanny-state censorship.”
Artificial intelligence may eventually be the solution, but for now Facebook has very little control over the content that gets broadcast on Facebook Live.