Facebook’s hate speech detection systems to focus on ‘worst of the worst,’ report says

Facebook is reportedly working on a major overhaul of its algorithms that police hate speech on its platforms. The changes, which are still in the early stages, include reenginering its automated moderation systems to better detect and remove hateful language that’s considered the “worst of the worst,” according to The Washington Post, citing internal Facebook documents. 

The massive social network has long banned hate speech, but its updated approach will shift the company away from treating all hate speech the same, the Post reported on Thursday. As part of the overhaul, Facebook systems will prioritize detecting and removing hateful comments directed at Black people, Muslims, people of more than one race, the LGBTQ community and Jews, the documents reportedly show. The system changes also reportedly include deprioritizing hateful comments against white people, Americans and men.

Facebook didn’t immediately respond to a request for comment, but told the Post that its systems are focused on finding hate speech that is the “most serious.”

“We know that hate speech targeted towards underrepresented groups can be the most harmful, which is why we have focused our technology on finding the hate speech that users and experts tell us is the most serious,” Facebook spokeswoman Sally Aldous told the Post. 

Facebook uses a mix of human reviewers and technology to remove harmful content. In November, the social network said its tools proactively detected 94.7% of the hate speech removed by the company between July to September. In the third quarter, Facebook took action against 22.1 million pieces of content for hate speech.

Despite its efforts, Facebook has been under fire from civil rights activists and politicians who say Facebook isn’t enforcing its rules against speech that directly attacks a person based on race, gender or other protected characteristics. Major brands this year paused spending on Facebook ads to pressure the company to do more to tackle hate speech, which they say is still slipping through on the social network.

The social network in November said there are 10 to 11 views of hate speech out of every 10,000 views of Facebook content. 

The first phase of the hate speech overhaul, apparently known as the WoW Project, was announced to a small group of Facebook employees in October, according to the Post. The overhaul was informed by extensive internal and external research, Aldous reportedly told the Post, as well as employees who have critiqued Facebook’s efforts to moderate hate speech.