Meta's Facebook platform failed to prevent the spread of anti-Rohingya hate speech in Myanmar, which contributed to violence against the Rohingya Muslim minority including killings, mass displacement, and ethnic cleansing.
Between 2013-2018, Meta's Facebook platform was used extensively in Myanmar to spread hate speech and misinformation targeting the Rohingya Muslim minority. The platform became the primary internet access point for Myanmar's 18 million users after telecommunications deregulation in 2013. Facebook had minimal content moderation resources for Burmese language content, with only 2 Burmese speakers reviewing content in 2015 and around 60 by 2018. The company outsourced monitoring to Accenture's 'Project Honey Badger' operation in Kuala Lumpur. Reuters found over 1,000 examples of anti-Rohingya posts calling Muslims 'dogs, maggots and rapists' and urging they 'be shot or exterminated.' UN investigators determined Facebook played a key role in spreading hate speech that fueled violence. The Myanmar military systematically used fake accounts and troll operations with up to 700 personnel to spread anti-Rohingya propaganda. This contributed to a military crackdown in August 2017 that forced over 730,000 Rohingya to flee to Bangladesh. Human rights groups documented killings, mass rape, and village burning. Facebook eventually removed some military accounts in August 2018 but was criticized as being 'too slow' to act.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
AI that exposes users to harmful, abusive, unsafe or inappropriate content. May involve providing advice or encouraging action. Examples of toxic content include hate speech, violence, extremism, illegal acts, or child sexual abuse material, as well as content that violates community norms such as profanity, inflammatory political speech, or pornography.
AI system
Due to a decision or action made by an AI system
Unintentional
Due to an unexpected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed