Facebook Goes to Extremes on Hate Speech

Facebook Goes to Extremes on Hate Speech
Facebook Goes to Extremes on Hate Speech

An accidental site-wide rollout of a new prospective feature on Facebook shows that the site is planning to give users the option to label posts and images “hate speech.” The site plugin is still in testing, but was accidentally added to every post on the site this week, giving people the option to tag everything from pictures of babies to posts about visits with Grandma as “hate speech.”

Key Facts

• The hate speech indicator inserts a subcaption on posts and images asking, “Does this post contain hate speech?” Users are then given the option to choose yes or no. There is nowhere to add context or explanations.
• It isn’t clear whether Facebook will ban based on responses alone or simply add a graphic content filter as they do with other posts. The first would be the harsher approach, while the second would leave posts and profiles intact while forcing readers to actively select whether they want to see the content.
• Facebook has come under serious fire lately for allowing true hate speech to flourish on the site unchecked. Many blame this on lax and algorithm-driven TOS policies that are essentially ineffective, penalizing the wrong people while allowing videos (like a beheading in Mexico) to circulate unchecked.
• However, some users are pointing out the fact that such a feature has a high propensity for misuse. The new hate speech indicator could be abused maliciously to produce suspensions or even bans, something that happens even now with the current reporting system.
• These concerns are even more valid now that it appears Facebook doesn’t understand how to properly balance the algorithm, including how and when it appears. Most similar algorithms work by detecting the content in a post; if it’s already selecting puppies and kittens, what else will it inappropriately select?
• If the site plugin remains intact in its current form, you can almost bet it will be misused by people to stifle free speech. This includes shutting down politicians, attacking activists, and potentially, engaging in social media witch hunts against people with unpopular opinions.