We’ve spent the last few years building tools, teams and technologies to help protect elections from interference, prevent misinformation from spreading on our apps and keep people safe from harmful content.
So when the COVID-19 crisis emerged, we had the tools and processes in place to move quickly and we were able to continue finding and removing content that violates our policies. When we temporarily sent our content reviewers home due to the COVID-19 pandemic, we increased our reliance on these automated systems and prioritized high-severity content for our teams to review in order to continue to keep our apps safe during this time.
This report includes data only through March 2020 so it does not reflect the full impact of the changes we made during the pandemic. We anticipate we’ll see the impact of those changes in our next report, and possibly beyond, and we will be transparent about them. For example, for the past seven weeks we couldn’t always offer the option to appeal content decisions and account removals, so we expect the number of appeals to be much lower in our next report. We also prioritized removing harmful content over measuring our efforts, so we may not be able to calculate the prevalence of violating content during this time. Today’s report shows the impact of advancements we’ve made in the technology we use to proactively find and remove violating content.
What’s New in This Report?
We are now including metrics across twelve policies on Facebook and metrics across ten policies on Instagram. The report introduces Instagram data in four issue areas: Hate Speech, Adult Nudity and Sexual Activity, Violent and Graphic Content, and Bullying and Harassment. For the first time, we are also sharing data on the number of appeals people make on content we’ve taken action against on Instagram, and the number of decisions we overturn either based on those appeals or when we identify the issue ourselves. We’ve also added data on our efforts to combat organized hate on Facebook and Instagram. You can learn more about these efforts and the progress we’ve made here.
Progress in Finding and Removing Violating Content
We improved our technology that proactively finds violating content, which helped us remove more violating content so fewer people saw it.
· On Facebook, we continued to expand our proactive detection technology for hate speech to more languages, and improved our existing detection systems. Our proactive detection rate for hate speech increased by more than 8 points over the past two quarters totaling almost a 20-point increase in just one year. As a result, we are able to find more content and can now detect almost 90% of the content we remove before anyone reports it to us. In addition, thanks to other improvements we made to our detection technology, we doubled the amount of drug content we removed in Q4 2019, removing 8.8 million pieces of content.
· On Instagram, we made improvements to our text and image matching technology to help us find more suicide and self-injury content. As a result, we increased the amount of content we took action on by 40% and increased our proactive detection rate by more than 12 points since the last report. We also made progress in our work combating online bullying by introducing several new features to help people manage their experience and limit unwanted interactions, and we announced new Instagram controls today. We are sharing enforcement data for bullying on Instagram for the first time in this report, including taking action on 1.5 million pieces of content in both Q4 2019 and Q1 2020.
· Lastly, improvements to our technology for finding and removing content similar to existing violations in our databases helped us take down more child nudity and sexual exploitative content on Facebook and Instagram.
Over the last six months, we’ve started to use technology more to prioritize content for our teams to review based on factors like virality and severity among others. Going forward, we plan to leverage technology to also take action on content, including removing more posts automatically. This will enable our content reviewers to focus their time on other types of content where more nuance and context are needed to make a decision.
The Community Standards Enforcement Report is published in conjunction with our bi-annual Transparency Report that shares numbers on government requests for user data, content restrictions based on local law, intellectual property take-downs and internet disruptions.