Facebook plans regional check for clean content

NEW DELHI: Facebook may look at region-specific content regulations to curb hate speech and fake news on its platform, as the social media company seeks to be more sensitive to the individual needs of countries and communities on such issues.

It is “challenging” to have one global standard as Facebook currently has, since what may be offensive to one community may not be so to another, said Guy Rosen, its vice president for product management.

“Sometime in the future there is room to think about something that is more maybe regionalised and something that does have a little more variation,” Rosen said in an interview to ET. At present, the company however is engaged to the fullest on enforcing the existing set of guidelines, he added.


According to global statistics released by Facebook on Tuesday, the company took down 2.45 million posts related to hate speech in the first quarter of 2018, up from 1.61 million in the last quarter of 2017. However, it was able to discover only 38% of the content before users flagged it, mostly through its technology — artificial intelligence and machine learning. It did not share country-specific data.

“Even Mark (Zuckerberg, Facebook founder) early last year in his letter to the community said that how people in different cultures will have different approaches. At some point we will have to evolve how we think about standards to be more sensitive to those varied needs,” said Rosen.

This is the first time Facebook has publicly released data on the number of posts it took down across various categories, giving also a comparison with the previous quarter. Acknowledging the mammoth fake accounts problem, Facebook said it detected that 3-4% of the total monthly active users on the platform were using fake accounts. The percentage was consistent in the first quarter of 2018 with a total 583 million accounts. Facebook claimed that it was able to detect 98.5% of the fake accounts soon after they were created in the first three months of 2018. It did not share the exact data for how prevalent was terrorist propaganda violations on Facebook, but said it took down 1.86 million posts related to terror in Q1, compared with 1.08 million in the last quarter of 2017.

Facebook released the global numbers in the first edition and Rosen said the company is working on releasing numbers which are country or region specific. “It is definitely something that we want to do. We want to get there. There is a bunch of things we need to figure out,” he said. “People like spammers, for example, they pretend and they connect through different ways. They pretend to be from different countries. How do we map that? There is also the issue of a person who produced the content and the person through whom it was routed. So, we have to take a bunch of decisions and understand how to categorize it...”

He said even anecdotally he was currently unable to give a sense of the issues concerning India versus markets like the US. India is the largest market for the company in terms of users. It has over 241 million users in India.