Facebook has released its latest progress reports on taking down harmful content from its social network and Instagram. The data shows a sharp increase in the number of pieces of hate speech taken down in the first quarter of 2020, to around 4.7 million from 1.6 million in the previous quarter, Facebook said.
A year since the Christchurch Call, when social media companies agreed to do more to halt the spread of hate speech and extremist content, Facebook issued a statement highlighting its efforts to take down the harmful content more quickly. The company said it was improving constantly its automatic detection techniques to identify organised hate on its apps, and nearly 97 percent of the items removed in Q1 2020 were detected before it was alerted by users. The figure was lower at around 70 percent on Instagram, but the number of items removed also was much lower, at 175,000 pieces of hate speech.
What Facebook calls 'organised hate' content focuses on terrorist propaganda and other hate organizations that engage in coordinated violence against others based on characteristics such as religion, race, ethnicity or national origin. This is only one part of the harmful content it aims to remove from the social network, as its latest Community Enforcement Standards report shows. It also removed around 25.5 million posts involving violent acts in Q1, 8.6 million posts with child nudity or secural exploitation, and 8.1 million posts involving adult nudity or sexual activity.
The company also released its latest transparency report on government requests for user data. These increased by 9.5 percent in the second half of 2019 to almost 141,000. Of the total volume, the US continues to submit the largest number of requests (51,121), followed by India, the UK, Germany and France. The report shows it also removed nearly 5 million pieces of content in the six months for reports of intellectual property infringement.
Facebook's latest reports also show it disabled some 1.7 billion fake accounts on the site in the first quarter, up from 1.1 billion in Q4 2019. That's the same as the number of daily active users reported earlier by the company for Q1. The company estimates around 5 percent of its monthly active user base is so-called fake accounts. It detects over 99 percent of the accounts removed before receiving any alert from users.
