Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Memory Power Asymmetry in Human-AI Relationships: Preserving Mutual Forgetting in the Digital Age

Rasam Dorri, Rami Zwick

TL;DR

AI-enabled memory creates a structural power imbalance in human–AI relationships, termed Memory Power Asymmetry (MPA). The paper defines MPA across four dimensions (persistence, accuracy, accessibility, integration), identifies four mechanisms driving power (strategic deployment, narrative control, dependence, vulnerability), and elaborates multi-level consequences with boundary propositions. It differentiates MPA from information asymmetry, privacy, surveillance, and CRM, and proposes six design principles—forgetting-by-design, memory transparency, goal alignment, augmentation, contextual containment, and ecosystem oversight—to restore balance. The work offers a forward-looking research and policy agenda to safeguard autonomy, fairness, and social mobility in memory-rich AI ecosystems.

Abstract

As artificial intelligence (AI) becomes embedded in personal and professional relationships, a new kind of power imbalance emerges from asymmetric memory capabilities. Human relationships have historically relied on mutual forgetting, the natural tendency for both parties to forget details over time, as a foundation for psychological safety, forgiveness, and identity change. By contrast, AI systems can record, store, and recombine interaction histories at scale, often indefinitely. We introduce Memory Power Asymmetry (MPA): a structural power imbalance that arises when one relationship partner (typically an AI-enabled firm) possesses a substantially superior capacity to record, retain, retrieve, and integrate the shared history of the relationship, and can selectively deploy that history in ways the other partner (the human) cannot. Drawing on research in human memory, power-dependence theory, AI architecture, and consumer vulnerability, we develop a conceptual framework with four dimensions of MPA (persistence, accuracy, accessibility, integration) and four mechanisms by which memory asymmetry is translated into power (strategic memory deployment, narrative control, dependence asymmetry, vulnerability accumulation). We theorize downstream consequences at individual, relational/firm, and societal levels, formulate boundary-conditioned propositions, and articulate six design principles for restoring a healthier balance of memory in human-AI relationships (e.g., forgetting by design, contextual containment, symmetric access to records). Our analysis positions MPA as a distinct construct relative to information asymmetry, privacy, surveillance, and customer relationship management, and argues that protecting mutual forgetting, or at least mutual control over memory, should become a central design and policy goal in the AI age.

Memory Power Asymmetry in Human-AI Relationships: Preserving Mutual Forgetting in the Digital Age

TL;DR

AI-enabled memory creates a structural power imbalance in human–AI relationships, termed Memory Power Asymmetry (MPA). The paper defines MPA across four dimensions (persistence, accuracy, accessibility, integration), identifies four mechanisms driving power (strategic deployment, narrative control, dependence, vulnerability), and elaborates multi-level consequences with boundary propositions. It differentiates MPA from information asymmetry, privacy, surveillance, and CRM, and proposes six design principles—forgetting-by-design, memory transparency, goal alignment, augmentation, contextual containment, and ecosystem oversight—to restore balance. The work offers a forward-looking research and policy agenda to safeguard autonomy, fairness, and social mobility in memory-rich AI ecosystems.

Abstract

As artificial intelligence (AI) becomes embedded in personal and professional relationships, a new kind of power imbalance emerges from asymmetric memory capabilities. Human relationships have historically relied on mutual forgetting, the natural tendency for both parties to forget details over time, as a foundation for psychological safety, forgiveness, and identity change. By contrast, AI systems can record, store, and recombine interaction histories at scale, often indefinitely. We introduce Memory Power Asymmetry (MPA): a structural power imbalance that arises when one relationship partner (typically an AI-enabled firm) possesses a substantially superior capacity to record, retain, retrieve, and integrate the shared history of the relationship, and can selectively deploy that history in ways the other partner (the human) cannot. Drawing on research in human memory, power-dependence theory, AI architecture, and consumer vulnerability, we develop a conceptual framework with four dimensions of MPA (persistence, accuracy, accessibility, integration) and four mechanisms by which memory asymmetry is translated into power (strategic memory deployment, narrative control, dependence asymmetry, vulnerability accumulation). We theorize downstream consequences at individual, relational/firm, and societal levels, formulate boundary-conditioned propositions, and articulate six design principles for restoring a healthier balance of memory in human-AI relationships (e.g., forgetting by design, contextual containment, symmetric access to records). Our analysis positions MPA as a distinct construct relative to information asymmetry, privacy, surveillance, and customer relationship management, and argues that protecting mutual forgetting, or at least mutual control over memory, should become a central design and policy goal in the AI age.

Paper Structure

This paper contains 46 sections, 2 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Mutual Forgetting and Divergent Memory Systems
  3. Mutual forgetting as a norm in human relationships
  4. Human memory: episodic and semantic systems
  5. AI memory architectures: parametric, non-parametric, and nested systems
  6. Defining Memory Power Asymmetry (MPA)
  7. Boundary conditions and moderators of MPA
  8. How MPA Differs from Related Constructs
  9. Information asymmetry
  10. Privacy and data asymmetry
  11. Surveillance and surveillance capitalism
  12. Customer relationship management (CRM)
  13. Episodic vs. semantic memory as core distinction
  14. Mechanisms: How Asymmetrical Memory Translates to Power
  15. Strategic memory deployment
  16. ...and 31 more sections

Figures (2)

  • Figure 1: From memory asymmetry to power: mappings from MPA dimensions (persistence, accuracy, accessibility, integration) to mechanisms (strategic memory deployment, narrative control, dependence asymmetry, vulnerability accumulation); solid arrows denote primary relationships, dashed arrows supporting relationships.
  • Figure 2: Mitigations to rebalance memory power: six principles, goal-aligned memory; memory transparency (visibility, interpretability, contestability); forgetting-by-design (retention limits/decay, amnesty/reset controls, friction for old/high-risk reuse); augmentation (support, not replace); contextual containment (domain silos, purpose limitation, explicit consent); ecosystem safeguards (retention caps, portability/erasure/unlearning, independent audits, public reporting), with targeted MPA dimensions noted.