ARTIFICIAL INTELLIGENCE-ASSISTED DECISION SYSTEMS FOR BROKERAGE OPERATIONS AND CAPITAL MARKET RISK MANAGEMENT IN SAUDI ARABIA

Artificial intelligence is moving from experimental analytics to operational decision support across brokerage, trading, surveillance, and market-risk functions. In Saudi Arabia, this transition intersects with a rapidly modernizing capital market,expanding digital participation, and a policy environment that centers market development and technological capability in economic diversification. This review critically evaluates how artificial intelligence-assisted decision systems can be designed, governed, and applied to brokerage operations and capital market risk management in the Saudi context. An integrative review approach synthesizes peer-reviewed research on machine learning, deep learning, reinforcement learning, explainable artificial intelligence, asset pricing, forecasting, algorithmic trading, portfolio management, and financial risk. The evidence indicates that artificial intelligence can improve signal extraction, execution support, client segmentation, portfolio construction, volatility forecasting, and surveillance, but predictive gains do not automatically translate into robust economic or regulatory outcomes. Model instability, regime shifts, transaction costs, data leakage, limited explainability, and weak human oversight can reverse apparent benefits. Saudi-specific studies demonstrate promising applications to Tadawul forecasting and automated trading, yet the literature remains fragmented and concentrated on prediction rather than end-to-end brokerage governance.