The Riots Facebook Later Apologized For
In early 2018, anti-Muslim mobs attacked homes, shops and mosques across Sri Lanka's Kandy district after inflammatory posts and rumors spread on Facebook. The government declared a state of emergency and blocked the platform nationwide, and Facebook later apologized for its role.
The unrest began in late February 2018 in the eastern town of Ampara, where a false rumor circulated on Facebook that a Muslim restaurateur had mixed sterilization pills into food served to Sinhalese Buddhist customers, according to contemporaneous accounts. Days later the violence spread to the central Kandy district, where mobs burned Muslim-owned homes and businesses from March 2 to March 10.
On March 6, the government declared a nationwide state of emergency, its first since 2011, and imposed a curfew in the affected areas, CNN reported. At least two people were killed and mosques, homes and businesses were destroyed in the Kandy district, Al Jazeera reported.
The platform is switched off
As the violence spread, authorities took the unusual step of blocking Facebook, WhatsApp, Instagram and other services across the country to stop mobs from organizing and to slow the spread of rumors, according to Al Jazeera, which noted that Facebook had more than 5.5 million users in Sri Lanka at the time. Government officials openly linked the platform to the bloodshed.
“Facebook does not have adequate resources to monitor posts in Sinhalese,” Harin Fernando, the country's telecommunications minister, told Al Jazeera as the ban took hold. Officials said the company had only a handful of people reviewing content in Sinhala, the language of the country's ethnic majority, whose members were behind much of the violence, Al Jazeera later reported.
Warnings, then an apology
A New York Times investigation published weeks after the riots reconstructed the descent from rumor to killing and concluded that Facebook's news feed had played a central role, drawing on interviews with officials, victims and users, the report found. Civil-society groups in Sri Lanka had warned the company for years that hateful content on its platform could spill into real-world violence.
Two years later, Facebook commissioned an independent assessment of its role. The review found that hate speech and rumors on the platform “may have led to” offline violence and that the company had failed to remove such content before the unrest, according to Al Jazeera. In a statement responding to the findings, the company said: “We deplore the misuse of our platform. We recognise, and apologise for, the very real human rights impacts that resulted,” Al Jazeera reported. The company said it had since hired more staff, including Sinhala speakers, and expanded automated detection.
A recurring pattern
Analysts have described the Sri Lankan case as part of a broader pattern in which platforms invest little in moderating content in smaller languages, leaving rumors and incitement to circulate largely unchecked until violence erupts, The Diplomat noted. The episode remains one of the earliest widely documented instances in which a government switched off a major social platform to try to halt communal violence.