The expansion of digital services, such as streaming, cloud storage, and blockchain, has resulted in a large increase in energy usage in data centers, networks, and end-user devices, which contributes considerably to global greenhouse gas emissions. Despite this consequence, end users are largely ignorant of their personal digital carbon footprint. In this study, we provide EcoCarbon, a lightweight, AI-driven monitoring platform that estimates session-level energy consumption and associated carbon emissions for common digital activities such as online surfing, video streaming, file transfers, and device charging. EcoCarbon uses embedded usage-tracking agents and a mobile dashboard to deliver real-time, individualized carbon effect feedback using a supervised Artificial Neural Network (ANN) trained on vast historical power use profiles. An explainable AI module converts these predictions into actionable recommendations, such as improving video quality, scheduling downloads during off-peak hours, or managing background activities, all with the goal of encouraging sustainable digital behavior. EcoCarbon is a scalable solution for promoting greener digital practices and enhancing digital sustainability by combining fine-grained monitoring, precise AI-based estimation, and user-tailored behavioral guidance.