Authors: Dr. Harsha Sammangi, Aditya Jagatha, Navyasri Maddukuri
Abstract: Climate disruption has become a persistent organizational condition rather than an episodic event, yet most information systems designed to support organizational resilience treat each disruption as an isolated incident. Existing digital resilience platforms, disaster recovery systems, and AI-driven decision support tools lack the capacity to accumulate, encode, and reuse organizational knowledge across successive climate-related crises. This paper introduces Algorithmic Resilience Memory (ARM), a novel IS construct defined as an AI-enabled organizational capability through which agentic AI systems sense climate-related disruptions, encode prior organizational responses, preserve decision rationale, generate contextually adaptive recommendations, and reconfigure future actions through structured outcome feedback. Drawing on Design Science Research (DSR), we propose and develop an Agentic AI-Based Algorithmic Resilience Memory Framework as the primary artifact. The framework integrates six interdependent functional layers—environmental sensing, knowledge encoding, agentic AI reasoning, explainable decision support, human governance, and adaptive learning—grounded in organizational memory theory, dynamic capabilities theory, sociotechnical systems theory, and responsible AI governance principles. We demonstrate the framework through a detailed scenario involving a regional flood disrupting a manufacturing firm's supply chain operations and evaluate its utility using scenario-based assessment and expert panel validation. The paper makes three primary contributions: it introduces ARM as a theoretically grounded IS construct that advances digital resilience research; it offers a design-science artifact that organizations can adopt for AI-driven climate-crisis adaptation; and it establishes design principles for building agentic AI systems capable of institutional learning across repeated climate disruptions.
DOI: http://doi.org/10.5281/zenodo.20676882
