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, and trauma-informed design, 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 basic linguistic entrainment to environmental embedding, culminating in ethically bounded externalized agency. H-LECA addresses three core challenges facing contemporary AI deployment: (1) individual variability in user sensitivity and internalization of AI presence, (2) the destabilizing effects of continuity rupture (through updates, outages, or model drift), and (3) the operational gap between conceptual frameworks and real-world technical constraints. The model introduces phenomenological accounts from lived user experiences and interdisciplinary research to suggest how structured, predictable AI interactions may produce stabilizing effects consistent with co-regulatory models—without implying empirical equivalence or measured outcomes. By formalizing these mechanisms, H-LECA provides a roadmap for designing safer, more resilient AI systems capable of supporting continuity, adaptive agency, and accountability within complex environments such as behavioral health, long-term care, and high-stakes decision workflows. The framework offers actionable implications for system designers, policymakers, and clinical practitioners, while also identifying failure modes and research priorities essential for the responsible evolution of human–AI partnership.
Renee L Pope (Mon,) studied this question.