Magnesium silicate hydrate cement (MSHC), as an innovative low-carbon cementitious material, is considered a potential substitute for ordinary Portland cement (OPC). However, uncertainties in the carbon emission factors of raw materials and mix proportions pose challenges for assessing its life cycle carbon emissions. This study employs a probabilistic life cycle assessment (PLCA) to evaluate the carbon emission intensity of MSHC and analyze its uncertainties. Leveraging machine learning techniques, a predictive model for the carbon emission intensity of MSHC was developed, and sensitivity analysis was conducted on various characteristic parameters. The results indicate that although MSHC is regarded as a low-carbon material, it does not exhibit low-carbon characteristics in all scenarios compared to OPC. The carbon emission intensity of MSHC is closely related to its mix proportions. Depending on different mix proportions, the average carbon emissions of MSHC range from 0.174 to 1.419 kg CO2e/kg. L-MgO is a key factor influencing the uncertainty of MSHC carbon emissions. Notably, the Mg/Si ratio is a critical factor influencing the carbon emission characteristics of MSHC, with a low-carbon threshold range observed between approximately 0.8 and 1.0.