Facebook says it is trying to do a better job of finding and removing terrorist content, and on Thursday the world's largest social network said it has seen progress .
Facebook said it took down 3 million posts about terrorismdown from the 9.4 million posts it removed in the second quarter. The average time terrorist content remained on the platform after users flagged it fell from 43 hours in the first quarter to 18 hours in the third quarter, the company said.
The social networks are under pressure to remove terrorist content before the spread of violence in the real world. However, as they increase their efforts, the perpetrators constantly change their strategy to avoid detection, say the companies. Some terrorists try to create new accounts or encrypt their messages, Facebook said.
“We can reduce the presence of terrorism on key social platforms, but completely eliminating it requires addressing the people and organizations that produce this material in the real world,” said Monika Bickert, Facebook’s global head of policy management, and Brian Fishman, the company’s head of counterterrorism.
Facebook relies on machine learning to detect terrorist content and heavily relies on the people who report inappropriate content. Sometimes the company will automatically download certain posts if the system considers the user marking it as terrorist to be “highly trusted”. The company has also expanded some of its tools into more languages.
The third quarter of the year, about 99% of the content related to ISIS and al-Qaeda was taken down by technology before a user needs to report it to Facebook.
