Overexposure to stress can lead to serious physiological and psychological issues, highlighting the importance of stress management for long-term well-being. The aim of this work is to create a stress recognition system using artificial intelligence (AI), in response to the increasing demand for accurate stress detection. The system uses data from wearable sensors placed on the chest and wrist. It incorporates a neural network architecture with an autoencoder to extract features from the data. These features capture important properties for stress classification tasks. The system’s performance was evaluated for both 2-class (stress, non-stress) and 3-class (baseline, stress, amusement) stress classification using several performance metrics. The proposed approach achieved an accuracy of 96% in 2-class classification and 90% in multi-class classification. This study demonstrates the potential of integrating AI with wearable technologies to effectively detect stress. Integrating AI-driven stress detection with wearable technology not only improves personal health management but also promotes sustainable practices in healthcare and business wellness programs, hence enhancing overall well-being and productivity.