The integration of artificial intelligence (AI) technologies into the criminal justice system is progressing rapidly, often outpacing our comprehensive understanding of its legal, ethical, and social ramifications. While AI offers significant potential for enhancing procedural efficiency, its deployment necessitates a critical assessment of its impact on fundamental principles of justice. This study aims to explore the dual impact of AI tools—specifically in risk assessments and sentencing practices—by examining their capacity to streamline legal procedures alongside the challenges they introduce concerning fairness, transparency, and accountability. Employing a mixed-methods approach, we analyze quantitative data from 570 instances of AI usage in case processing and qualitative data gathered through interviews with 40 legal professionals. Our analysis focuses on three key variables—Sentencing Time Deviation, Risk Assessment Bias Coefficient, and Regulatory Compliance Index—to empirically test the hypothesis that AI enhances efficiency while simultaneously raising significant ethical and legal concerns. The findings confirm a statistically significant reduction in processing times but also reveal persistent, quantifiable biases in risk assessment outcomes, particularly across demographic groups. Furthermore, a comparative review of regulatory frameworks across jurisdictions highlights the urgent need for standardized, transparent governance structures to ensure that AI-driven systems promote equitable and accountable justice.