Dive Brief:

• An hour-long video of a man being violently assaulted was broadcast via Facebook Live Wednesday, according to a report in Ad Age. Chicago police have arrested and charged four suspects in relation to what's being deemed a hate crime, per ABC News.
• As Ad Age points out, marketers and publishers may be more wary of using Facebook Live following the incident for fear of being associated with extreme videos. Facebook Live and live streaming, in general, are particularly risky formats, as content appears in real-time as opposed to being edited or censored beforehand.
• "I don't think the incident will intimidate brands who want to move into live broadcasting," Jill Sherman, the head of social media at the agency DigitasLBi, told Ad Age. "That said, brands should absolutely be thinking about how they place media in social spaces to avoid having their content directly aligned with the wave of content and commentary that follow these types of incidents."

Dive Insight:

Live streaming is still a nascent format and the horrific torture video sheds light on the pitfalls platforms like Facebook Live can fall into. The fact that the stream reportedly broadcast for an hour without getting taken down might suggest a lack of moderation tools for flagging and removing extreme content from user feeds.  

"We do not allow people to celebrate or glorify crimes on Facebook and have removed the original video for this reason," a Facebook spokesperson said in a statement to Ad Age. "In many instances, though, when people share this type of content, they are doing so to condemn violence or raise awareness about it. In that case, the video would be allowed."

Brands and their media buyers obviously don’t want to appear to be supporting videos that depict torture either, but the real-time aspect of streaming makes moderating things like ad placement more difficult than on more traditional or static formats. Facebook Live introduced mid-roll ads for publisher streams in August last year. While a user stream depicting torture is different from an official publisher broadcast, there’s the possibility that a mid-roll ad could play during a publishers’ stream covering similarly disturbing news or imagery.  

Facebook is taking several steps to try and account for extreme content posted to its platform. Facebook Mentions, a live streaming tool for celebrities, politicians and other notable public figures, introduced the ability to include pre-broadcast prompts and comment moderation in December.

Reports last month suggested that the social network was developing a machine learning tool that removes nudity, violence and other Facebook Live content not in line with site policies. While AI might be able to flag down unseemly streams more efficiently than a human task force, it doesn’t necessarily possess the judgment skills to gauge what’s actually appropriate or inappropriate in context (for example, historical documents or art featuring nudity/violence could be acceptable).