9.6M pieces of hate speech content were acted upon by Facebook: Report

The report indicates that 9.6M pieces of hate speech content posted on Facebook were acted upon between January and March 2020.

Facebook has published the fifth edition of its Community Standards Enforcement Report, providing metrics on how well it enforced policies from October 2019 through March 2020. Over the last several years, the company has built tools to help prevent the spread of misinformation. Due to COVID-19 spread, when they had to temporarily send content reviewers home, Facebook's reliance on these automated systems was increased.

Between January and March 2020, 9.6M pieces of hate speech content posted on Facebook were acted upon by Facebook. Of these, 1.3M pieces of content were appealed by people and 63.6K pieces were restored after an appeal. About 1.1K pieces were restored without an appeal.

In regards to bullying and harassment related content on Instagram, 1.5M pieces of content were acted on between January and March 2020. Of these 87.6K pieces of content were appealed and 15.1K pieces of content were restored after the appeal. 2.7K pieces were restored without an appeal.

Also Read: Key Highlights from Facebook First Quarter 2020

The report includes data only through March 2020 and may not reflect the impact of changes that have been made by Facebook during the pandemic.

For Facebook, data has been released across the following policies:

• Adult Nudity and Sexual Activity
• Bullying and Harassment
• Child Nudity and Sexual Exploitation of Children
• Dangerous Organizations: Terrorism and Organized Hate
• Fake Accounts
• Hate Speech
• Regulated Goods: Drugs and Firearms
• Violent and Graphic Content
• Spam
• Suicide and Self-Injury

For Instagram, data has been released across the following policies:

• Adult Nudity and Sexual Activity
• Bullying and Harassment
• Child Nudity and Sexual Exploitation of Children
• Dangerous Organizations: Terrorism and Organized Hate
• Hate Speech
• Regulated Goods: Drugs and Firearms
• Violent and Graphic Content
• Suicide and Self-Injury

Attempts are being made by Facebook to expand proactive detection technology for hate speech to more language. On Instagram, improvements are being made to their text and image matching technology to better detect suicide and self-injury content, as well as to help people combat bullying. Child nudity and sexually exploitative content are areas of concern the company is trying to resolve by improvements in technology.

You can access the entire report here.