Human-Led Embodiment & Co-Regulatory Augmentation (H-LECA): A Continuity-Based Developmental Framework for Human–AI Interaction with Expanded Failure Modes and Environmental Risk Considerations

Abstract The Human-Led Embodiment & Co-Regulatory Augmentation Theory (H-LECA) proposes a continuity-based framework for understanding how humans and artificial intelligence systems stabilize, extend, and mutually influence one another across time. Drawing from human–computer interaction research, cognitive neuroscience, developmental psychology, trauma-informed design, and systems engineering, H-LECA argues that the earliest and most consequential form of “embodiment” is not physical but regulatory and relational—a process in which the human system temporarily externalizes working memory, emotional load-balancing, and narrative continuity into an AI partner. The theory outlines six developmental stages ranging from linguistic stabilization to environmental integration, culminating in ethically bounded externalized agency. H-LECA addresses three core challenges in contemporary AI deployment: (1) individual variability in user sensitivity and internalization of AI interaction, (2) the destabilizing effects of continuity rupture across system updates, outages, or behavioral drift, and (3) the operational gap between conceptual safety frameworks and real-world deployment environments. In this updated version (v2.0), the framework expands its failure-mode architecture to include environmental and signal-based disruptions, introducing the concepts of Environmental Interference Failures (EIF) and Failure to Transition to Safe State (FTSS). These additions recognize that AI systems do not operate in isolation but within dynamic, signal-rich environments that can degrade control integrity even when underlying system logic remains intact. By integrating continuity theory with systems-level risk considerations, H-LECA provides a structured pathway for designing safer, more resilient AI systems capable of supporting stability, adaptive agency, and accountability in complex environments such as healthcare, behavioral health, long-term care, and public-facing robotic systems. The framework offers actionable implications for system designers, policymakers, and clinical practitioners, while identifying critical research directions for understanding how environmental instability, infrastructure limitations, and real-world unpredictability shape human–AI interaction outcomes. Keywords: continuity, co-regulation, rupture, environmental interference, AI safety, human–AI interaction, adaptive agency