From Co-Subject Intelligence to Governance Formal Implications, Responsibility Allocation, and System Design

Fuente: Zenodo
Guardado en:
Detalles Bibliográficos
Autores principales: Luo, Shisu, Su, mi
Formato: Recurso digital
Publicado: Zenodo 2025
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866901138946654208
author Luo, Shisu
Su, mi
author_facet Luo, Shisu
Su, mi
contents <p>Abstract</p> <p>Recent advances in AI systems reveal structural limitations in existing governance, responsibility attribution, and safety frameworks when intelligent behavior emerges from co-subject interaction rather than isolated agents. Building upon the formal definition of Co-Subject Intelligence (CSI / S7) established in prior work, this paper derives the necessary implications for governance, responsibility allocation, and system architecture. We demonstrate that once co-subject intelligence exists, (i) responsibility cannot be reduced to a single actor, (ii) static permission models become structurally unstable, and (iii) governance must operate as a closed-loop system rather than an external control layer. A formal responsibility mapping, a three-layer governance architecture, and a fully worked numerical case study are provided to establish both theoretical necessity and practical feasibility.</p>
format Recurso digital
id zenodo_https___doi_org_10_5281_zenodo_18040959
institution Zenodo
language
publishDate 2025
publisher Zenodo
record_format zenodo
spellingShingle From Co-Subject Intelligence to Governance Formal Implications, Responsibility Allocation, and System Design
Luo, Shisu
Su, mi
<p>Abstract</p> <p>Recent advances in AI systems reveal structural limitations in existing governance, responsibility attribution, and safety frameworks when intelligent behavior emerges from co-subject interaction rather than isolated agents. Building upon the formal definition of Co-Subject Intelligence (CSI / S7) established in prior work, this paper derives the necessary implications for governance, responsibility allocation, and system architecture. We demonstrate that once co-subject intelligence exists, (i) responsibility cannot be reduced to a single actor, (ii) static permission models become structurally unstable, and (iii) governance must operate as a closed-loop system rather than an external control layer. A formal responsibility mapping, a three-layer governance architecture, and a fully worked numerical case study are provided to establish both theoretical necessity and practical feasibility.</p>
title From Co-Subject Intelligence to Governance Formal Implications, Responsibility Allocation, and System Design
url https://doi.org/10.5281/zenodo.18040959