There are severe impacts and consequences to humans, societies, and the environment due to global warming. Though there are various activities that contributes to global warming, the major contributor is carbon dioxide (CO2) emissions. Human activities release large amounts of carbon dioxide from the burning of fossil fuels, such as oil, gas, or coal in producing energy. Net zero is the new ambition of industries in balancing the CO2emissions in environment. Thus, this study finds the best predictive model for CO2emissions using machine learning model with the dataset of CO2emissions from 1991 until 2020. Machine Learning techniques is an efficient approach to study the CO2emissions prediction and has been very appealing to few research. The dataset is split into a train-test (estimation-validation) set with 80% train set and 20% test set (80:20) proportion. The predictive model was developed using Random Forest, Support Vector Machine and Artificial Neural Network algorithms with different parameters to get the outcome. The predictive model's performance was evaluated based on the error measurement metric of Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE). Its reveals that Support Vector Machine with linear kernel function is the best model among others which produces 65.7254 Mean Absolute Error (MAE), 112.2196 Root Mean Square Error (RMSE) and 0.2279% Mean Absolute Percentage Error (MAPE) from the train set. For industries committed to net zero carbon emissions, this analysis will be an advising factor on the prediction system to find the CO2emissions and how much fossil fuels’ reduction is required in achieving net zero carbon emission by 2050.