EIPINOVA
EIPINOVA
  • Home
  • Publications
    • Articles
    • Working Papers
    • Reports
    • Policy Brief
    • White Paper
    • Index Methodology Paper
    • Conference Papers
    • Journal Articles
  • Centers
    • Global AI Governance
    • AI & Societal Evolution
    • AI & Emerging Tech
    • AI & Human Resilience
    • Maritime History & Tech
  • Showcase
    • MCCM v2.3
  • About Us
    • About Us
    • Contact US
  • More
    • Home
    • Publications
      • Articles
      • Working Papers
      • Reports
      • Policy Brief
      • White Paper
      • Index Methodology Paper
      • Conference Papers
      • Journal Articles
    • Centers
      • Global AI Governance
      • AI & Societal Evolution
      • AI & Emerging Tech
      • AI & Human Resilience
      • Maritime History & Tech
    • Showcase
      • MCCM v2.3
    • About Us
      • About Us
      • Contact US
  • Home
  • Publications
    • Articles
    • Working Papers
    • Reports
    • Policy Brief
    • White Paper
    • Index Methodology Paper
    • Conference Papers
    • Journal Articles
  • Centers
    • Global AI Governance
    • AI & Societal Evolution
    • AI & Emerging Tech
    • AI & Human Resilience
    • Maritime History & Tech
  • Showcase
    • MCCM v2.3
  • About Us
    • About Us
    • Contact US

Publications

Conference Papers

Conference Papers | EPINOVA-2026-CP-01

Multiverse Path Encryption for Securing Interpretation under Adversarial Access


This conference paper, published in the Proceedings of IMNS 2026 by IEEE, introduces Multiverse Path Encryption (MPE), an interpretation-control framework for post-compromise environments in which adversaries may observe ciphertext-derived outputs or queryable system states. Rather than mapping a ciphertext to a single plaintext, MPE uses key-conditioned traversal over a shared semantic graph to generate multiple coherent, query-consistent worlds. The paper formalizes world non-identifiability and decision uncertainty, explains how global coherence preserves structured ambiguity, and evaluates MPE against heuristic and adaptive querying attackers. Experimental results show sustained target-selection error, high path-reconstruction error, and substantial decision-ranking distortion, demonstrating that adversarial access does not necessarily produce reliable interpretation. 

Read More
Back to top

Copyright © 2025–2026 EPINOVA LLC

Email: contactus@epinova.org Phone: +1 678-667-8001

All Rights Reserved.

  • Home
  • Publications
  • About Us
  • Contact US
  • Privacy Policy

This website uses cookies.

We use cookies to analyze website traffic and optimize your website experience. By accepting our use of cookies, your data will be aggregated with all other user data.

Accept