2 Summary of Social Media Observation (2021-2023) Key Takeaways This summarizes three years of observation (from February 2021 to December 2023) on the ecosystem of political information on social media platforms, focusing on Facebook and Telegram. Using the framework our team has developed we were able to analyze 95,109 posts from 534 Facebook pages and 19,038 messages from 95 Telegram channels during the past three years (Note: During the initial phase, we included 6,045 posts from 171 Facebook accounts. However, due to limitations in accessing data from the account type, we only monitored the page type later.) However, only 15,250 Facebook posts and 1,241 Telegram messages contained content related to federalism, democracy, human rights, and conflict resolution. The following are key findings from our content analysis. R1 R1 R2 R3 R1 R2 Fig.1. Number of monitored Facebook posts and Telegram messages l l l Over the past three years, social media contents following the coup were primarily characterized by citizen mobilization (61 percent) focused on achieving more immediate goals. Citizens were basically called to participate in or support various non-violent activities such as protests, civil disobedience movements, or different campaigns against the coup or the Myanmar military. Along with the non-violent campaigns, social media also calls to support various military campaigns either against the Myanmar military or to suppress the democratic resistance alliance. We also found that there were contents promoting federalism, equality, social harmony, and citizens’ rights focusing on changing societal values for a longer-term goal. However, compared to the detailed, clear, and specific calls to action for citizen mobilization, these messages were more generalized and broader. Social media also contained content that increased division, distrust, and violence by targeting people based on their political views, gender, ethnicity, or religion.