Judicial decision-making involves complex reasoning based on legal facts, statutory interpretation, and precedent analysis. Although transformer-based language models have improved the automated analysis of legal documents, their opaque decision-making processes limit their reliability in sensitive judicial applications. The lack of transparency and systematic bias evaluation in existing legal AI systems raises important concerns regarding accountability, fairness, and ethical deployment. This study proposes a transformer-based explainable artificial intelligence framework for judicial verdict prediction using structured representations of legal documents. The framework combines contextual language modeling with attribution-based interpretability techniques to examine the reasoning process of the model. A novel metric, the Bias Attribution Index, is introduced to quantify the influence of sensitive attributes on model predictions. The proposed NEXAJudicia framework was systematically evaluated across two phases on Indian judicial case data. Phase-1, focused on bias classification using a bias-annotated subset, achieved strong performance (Accuracy: 93%, Macro-F1: 0.91, Weighted-F1: 0.93). Phase-2, addressing multi-class verdict prediction on 4,001 cases with an 80/20 split and 5-fold cross-validation, yielded robust results (Accuracy: 82.88%, Macro-F1: 0.64, Weighted-F1: 0.89). Attribution analysis further substantiates that model predictions are primarily influenced by legally relevant factual and contextual evidence rather than sensitive demographic attributes, thereby ensuring both interpretability and fairness.