Facebook says AI is aiding platform’s ability to remove hate speech

Facebook said Thursday artificial intelligence (AI) tools have helped the company crack down on and remove hate speech from the platform. 

The platform released information about hate speech and how it is acting to remove such content in a transparency report released Thursday, following mounting pressure about the tech giant’s handling of hate speech and bigotry. 

“Advancements in AI technologies have allowed us to remove more hate speech from Facebook over time, and find more of it before users report it to us,” Facebook’s report states. 

The company said it “proactively” detected about 95 percent of hate speech content it removed in the last three months. 

It’s a significant increase from the roughly 24 percent of hate speech Facebook proactively removed at the end of 2017, when it first began reporting metrics for hate speech, according to the transparency report. 

“Whether content is proactively detected or reported by users, we often use AI to take action on the straightforward cases and prioritize the more nuanced cases, where context needs to be considered, for our reviewers,” the report states. 

Facebook said it estimated the prevalence of hate speech on the platform from July 2020 to September 2020 was .10 percent to .11 percent, meaning out of every 10,000 views of content on Facebook, about 10 to 11 of them included hate speech. 

The company said it calculates hate speech prevalence by selecting a sample of content seen on Facebook and labeling how much it violates the platform’s hate speech policies. The social media giant said representative samples are sent to reviewers across different languages and regions, since hate speech “depends on language and cultural context.” 

Facebook uses a combination of user reports and technology to find hate speech on its platforms. The tech giant said it uses AI to “help prioritize content for review” so its team of viewers can focus on content that poses the most harm.  

Facebook has been under increasing scrutiny from advocacy groups and Democratic lawmakers over the handling of hate speech and misinformation. 

During a Senate Judiciary Committee hearing on Tuesday, Democrats pressed Facebook CEO Mark Zuckerberg on the platform’s handling of hate speech. Earlier in the week, a group of Democratic senators sent Facebook a letter urging the company to take further action against anti-Muslim bigotry spreading on its platform. 

Civil rights groups joined together over the summer to organize a Facebook advertising boycott to place pressure on the company to take action on hate speech. 

Facebook has moved in recent months to change some of its policies surrounding content moderation and removal under increasing pressure. 

Last month the company said it would start removing content denying the Holocaust and would tighten its ban on content spreading the QAnon conspiracy theory.