The sudden widespread menace created by the present global pandemic COVID-19\nhas had an unprecedented effect on our lives. Man-kind is going through\nhumongous fear and dependence on social media like never before. Fear\ninevitably leads to panic, speculations, and the spread of misinformation. Many\ngovernments have taken measures to curb the spread of such misinformation for\npublic well being. Besides global measures, to have effective outreach, systems\nfor demographically local languages have an important role to play in this\neffort. Towards this, we propose an approach to detect fake news about COVID-19\nearly on from social media, such as tweets, for multiple Indic-Languages\nbesides English. In addition, we also create an annotated dataset of Hindi and\nBengali tweet for fake news detection. We propose a BERT based model augmented\nwith additional relevant features extracted from Twitter to identify fake\ntweets. To expand our approach to multiple Indic languages, we resort to mBERT\nbased model which is fine-tuned over created dataset in Hindi and Bengali. We\nalso propose a zero-shot learning approach to alleviate the data scarcity issue\nfor such low resource languages. Through rigorous experiments, we show that our\napproach reaches around 89% F-Score in fake tweet detection which supercedes\nthe state-of-the-art (SOTA) results. Moreover, we establish the first benchmark\nfor two Indic-Languages, Hindi and Bengali. Using our annotated data, our model\nachieves about 79% F-Score in Hindi and 81% F-Score for Bengali Tweets. Our\nzero-shot model achieves about 81% F-Score in Hindi and 78% F-Score for Bengali\nTweets without any annotated data, which clearly indicates the efficacy of our\napproach.\n