<p>Early diagnosis of cervical cancer is a long-term problem, specifically, under the conditions of low-resource environment, access to sophisticated diagnostic tools is limited. In order to overcome this challenge, this paper proposes a hybrid surface plasmon resonance (SPR) biosensor where detection of the Squamous Cell Carcinoma Antigen (SCC-Ag) as an important biomarker of cervical cancer becomes label-free with the embedding of machine learning techniques. The sensor design is constructed following a multilayer meta surface design containing graphene, MoS₂ and MXenes which allow localization of the exciting electric fields and interaction with the biomolecules. Despite the fact the earlier versions have had a problem of signal reproducibility, complexity of fabrication, and real-time analysis, the limitations are overcome by implementation of intelligent modelling which uses 1D-Convolutional Neural Networks (1D-CNN) and analysis schemes. FDTD and TMM simulations allowed understanding and optimizing the proposed meta surface architecture, whereas AI models were then trained on sensor outputs and produced accurate prediction and classification. Findings indicate a low limit of detection of (0.02 ng/mL), a high value of sensitivity (186&#xa0;deg/RIU) and a classification accuracy higher than 97%. Robustness was proved by repeatability tests with minimal deviation. Compared to traditional and graphene sensors, the presented system has high sensitivity, resolution, and AI compatibility. Overall, the authors of the work will have created a new kind of biosensing platform integrating nanomaterials and artificial intelligence to allow manufacturing, real-time, and accurate detection of cervical cancer.</p> Graphical abstract <p></p>