By means of enhanced predictive analytics, intelligent automation, and real-time decision-making capacity, artificial intelligence (AI) has fundamentally changed Decision Support Systems (DSS). This work presents an AI-driven decision support system architecture combining generative artificial intelligence with reinforcement learning to improve accuracy, adaptability, and efficiency of decision-making. Using important performance parameters like precision, recall, and F1-score, the performance of artificial intelligence-based DSS is evaluated in a controlled experimental environment against conventional DSS models in financial risk management With a 15.9% increase in fraud detection accuracy, a 16.8% increase in recall, and a 16.7% improvement in F1-score comparing to conventional models, the empirical results show that AI-driven decision support systems show better accuracy. Wilcoxon Signed-Rank tests, paired t-tests, and ANOVA among other statistical tests confirm the value of improvements in artificial intelligence. This paper addresses ethical issues, workforce adaptation, and possible AI usage in Decision Support Systems (DSS) related concerns. Future studies aim to increase artificial intelligence explainability, offer scalable AI-DSS solutions across several industries, and apply dynamic learning techniques to support decision-making in evolving environments.