GAILC AI Liability Exposure Map (ALEM): Identifying Potential AI Liability Exposure Before Harm Occurs—An Eight-Domain Framework for Preventive Examination

The GAILC AI Liability Exposure Map (ALEM) is an eight-domain analytical framework developed by the Global AI Liability Consortium (GAILC) to support the preventive identification and investigation of potential AI liability exposure before harmful consequences occur. As artificial intelligence becomes integrated into consequential organizational decisions and autonomous workflows, existing governance arrangements may not fully resolve questions about decision authority, meaningful human intervention, reliance on AI-generated outputs, operational execution, evidentiary sufficiency, and responsibility across multiple actors. ALEM addresses these questions through eight interconnected exposure domains: Authority Exposure Intervention Exposure Reliance Exposure Execution Exposure Allocation Exposure Evidence Exposure Consequence and Harm Exposure Change and Autonomy Exposure The working paper establishes domain definitions, distinguishing diagnostic questions, classification boundaries, evidence requirements, and a qualitative triage procedure for prioritizing preventive examination. A constructed AI-assisted financial approval workflow demonstrates the application of all eight domains, including three illustrative triage dispositions: Immediate Examination, Priority Remediation, and Routine Monitoring. The framework draws upon established research concerning human–automation interaction, automation bias, effective oversight, and accountability. It also provides conceptual cross-references to the NIST AI Risk Management Framework, ISO/IEC 42001, the European Union Artificial Intelligence Act, and the United Kingdom Senior Managers and Certification Regime. Two practical appendices provide a preliminary assessment worksheet and a proposed assessor-consistency evaluation protocol. ALEM forms part of GAILC's Preventive AI Liability Readiness approach and connects to the GAILC Intervention Test, Allocation Test, AI System Liability Self-Check, and AI Liability Exposure Brief. The framework is presented as a foundational analytical contribution intended for practitioner application, independent examination, and future methodological refinement. Its classification reliability and predictive validity have not yet been independently established. Core principle: The map identifies conditions that warrant investigation. It does not declare that legal liability exists. Author: Deusdedit RuhangariyoInstitution: Global AI Liability Consortium (GAILC)Version: 1.1Publication year: 2026 Institutional mission: Interrogating AI Liability to Protect Organizations and the Public.