Jose Manuel Perez Rios

A Cybernetic Architecture for AI Governance: Integrating Identity, VSM, and Organizational Pathologies

This Festschrift chapter presents an organizational cybernetic approach that integrates the Viable System Model (VSM) with the Taxonomy of Organizational Pathologies (TOP) to address contemporary complexity and the diffusion of artificial intelligence (AI). It begins by positing “identity” as an essential dimension, alongside anatomy, physiology, and neurology, and requires analysts to adopt an external vantage point to clarify who the organization is, why it exists, and where its boundaries lie. From this identity-centred stance, environmental and organizational complexity are unfolded into recursion levels, guided by Ashby’s Law of Requisite Variety and structured through the Recursion Levels–Key Factors Matrix, enabling a coherent “multi-universe” cartography of PAN-VSM configurations. On this basis, the chapter describes a top down, four stage framework: clarification of identity, purpose, and boundaries; construction of the vertical recursive structure; diagnosis or design of each organization–environment pair using VSM components (Systems 1–5 and communication channels); and a coherence check across levels. This architecture directly addresses criticisms that VSM is static, outdated, or excessively objective by embedding dynamics through System 4, scenario modelling, and explicitly incorporating stakeholder perceptions through protocols for surfacing and integrating subjective views. TOP complements the framework with a controlled vocabulary of structural, functional, and information/communication pathologies that reveal flaws in recursion design, in the balance of Systems 1–5, and in critical channels and signals. In the context of AI diffusion, the combined VSM–TOP lens illuminates systemic risks such as (for example): ill-defined identities guiding AI strategies, institutional schizophrenia in conflicting AI policies, autopoietic “beasts” overpowering metasystems, or missing or distorted channels that amplify misinformation.