Downloads: 12
United States | Computer Science and Information Technology | Volume 14 Issue 4, April 2025 | Pages: 2112 - 2117
Causal Inference in Agentic AI: Bridging Explainability and Dynamic Decision Making
Abstract: This study investigated the essential incorporation of causal inference mechanisms into agentic AI systems to enhance explainability and improve dynamic decision-making capabilities. Although current agentic AI systems exhibit impressive autonomous functionalities, they primarily depend on correlation-based pattern matching, which restricts their ability to provide transparent explanations for decisions and effectively adapt to novel scenarios. Our research introduces a novel framework that integrates causal reasoning within agentic architectures, emphasizing how causal models bridge the gap between black-box decision-making processes and human-interpretable explanations. Through experimental evaluation across multiple domains, we demonstrate that causally aware agentic systems achieve significantly higher performance in dynamic decision environments, offer more actionable explanations, and generalize better to unseen scenarios than traditional approaches.
Keywords: Agentic AI, Causal Inference, Explainability, Dynamic Decision-Making, Large Language Models, Counterfactual Reasoning, Human-AI Collaboration, Decision Support Systems, XAI, Explainable AI
Received Comments
No approved comments available.