For many people in Myanmar (Burma), Facebook is the internet.
Hate speech has been a growing issue in Myanmar, particularly with increased access to the internet (and Facebook), and also since the outbreak of communal violence in Rakhine State in 2012.
Reuters and the Human Rights Center at UC Berkeley School of Law found more than 1,000 pieces of hate speech and misleading contents - including posts, images, and comments, were made to attack Myanmar Muslims on Facebook, with some had been posted as early as 2012.
One of the post published in December 2013, featured a picture of Rohingya-style food with the message "We must fight them the way Hitler did the Jews, damn kalars!" Another user commented on a blogpost depicting a boat full of Rohingya refugees arriving in Indonesia: “Pour fuel and set fire so that they can meet Allah faster."
Other posts used dehumanising language, used anti-Muslim images and others that fuel hatred.
This shows Facebook's effort to crack down hate speech in Myanmar, which has contributed to violent attacks against the minority Muslim population, have not been sufficient.
The social media company has faced warnings from many human rights groups and researchers.
They are blaming the platform for helping people to spread misinformation and promote hatred of Muslims, particularly the Rohingya, since 2013. The United Nations investigators have also condemned Facebook’s role as a vehicle for "acrimony, dissension and conflict."
Founder and CEO Mark Zuckerberg said that the company has hired 60 Burmese-speaking content moderators to review contents spread on the platform, with plans to add more when needed. But the company has been slow to respond to the growing crisis.
This comes as Facebook' user base in the country has exploded to 18 million (a third of the population), making hate speech easy to spread and difficult to control. In comparison, Facebook only had 1.1 percent of Myanmar using the platform back in 2012.
For many people in Myanmar, Facebook is one of the primary means for them to understand what's going on around them. The social media is also one of the first places to go to find news and entertainment, as well as messaging.
Facebook's growth in the country has been fueled by the fact that it is zero-rated by some of the country’s mobile phone operators, meaning that people aren't required to pay data charges to use it.
As for reporting hate speech, Facebook typically relies on its users to report it in the first place. But due to most Burmese websites render their fonts in a way that confuses Facebook's systems, the social media has had difficulties in interpreting Burmese text.
What's more, Facebook only started translating Burmese as of April 2018. Until that happened, users wanting to report a post have have had to do so in English.
In its attempts, Facebook has identified and removed several hate figures and groups from the platform, and also deleted pages linked to the monk-led nationalist group Ma Ba Tha, the Association for the Protection of Race and Religion.
After Reuters published its report, Facebook revealed that the company has proactively identified (not anymore relying on user reports) about 52 percent of the content it removed for hate speech in Myanmar. This is up from 13 percent in the last quarter of 2017.
To be able to do this, Facebook also relies on Artificial Intelligence to parse a high volume of posts and flag potential problems.
Zuckerberg acknowledged that AI has a long way to go before it can be perfected, saying that it will take five or 10 years to train AI to recognize the nuances.
The technology is being tested in Myanmar, where a low rate of posts are flagged by Facebook users for potential policy violations. The company is also taking similarly focused enforcement strategies in Sri Lanka, India, Cameroon and the Central African Republic.
"We have a responsibility to fight abuse on our products. This is especially true in countries like Myanmar where many people are using the internet for the first time, and Facebook can be used to spread hate and incite violence," said a Facebook spokeswoman. "It’s a problem we were too slow to spot – and why we’re now working hard to ensure we’re doing all we can to prevent the spread of misinformation and hate."
