AI
All aspects of AI: ethics, governance, and working papers
- AI Responsibility Escape: Mechanisms, Attribution Layers, and Control Architecture2026-09-18 · AI Ethics & Governance
It studies AI responsibility escape, subjectification narratives, black-box excuses, and attribution dilution, and argues that responsibility follows authorization.
- WP-AI-8.2 AI Responsibility Escape: Mechanisms, Attribution Layers, and Control Architecture2026-09-17 · AI Governance Working Papers
Installment 8.2 of the foundational AI safety and governance research series: it establishes the concept of "responsibility escape" and five layers of responsibility attribution (cognition, release, access, configuration, execution), with supporting control architecture including authorization records, capability-release registries, circuit-breaker events, and high-risk human sign-off.
- The Engineering Circuit Breaker2026-09-12 · AI Ethics & Governance
It proposes the engineering circuit breaker, dual-track governance, capability-boundary stop codes, and non-transferable release responsibility.
- WP-AI-8.1 Foundations of AI Safety and Governance: The Engineering Circuit Breaker2026-09-12 · AI Governance Working Papers
Installment 8.1 of the foundational AI safety and governance research series: centered on the "engineering circuit breaker," it argues for institutionally separating capability research from commercial deployment. Uncertainty triggers the breaker; critical junctures receive deterministic override and physical isolation; risk creators bear the burden of proof.
- From Atlas to a Billion Humanoid Robots: AI Safety Governance Is Approaching a Fundamental Redesign2026-09-03 · AI Ethics & Governance
World models and physical AI extend digital risks into physical feedback loops and irreversible cascades. AI safety governance requires fundamental redesign.
- The National AI Capability Series Is Now Published2026-08-24 · AI Ethics & Governance
The National AI Capability series is now published: it discusses AI strategies, governance policies, and development frameworks around the world.
- WP–AI–7 Series: Working Paper Series on AI Governance and Civilisational Foundations2026-07-25 · AI Governance Working Papers
Introduction to the WP-AI-7 series: Part I analyzes the structural blind spots of AI governance itself; Part II analyzes the deficit in AI governance's philosophical foundations.
- WP–AI–7 Working Paper Series on AI Governance and Civilisational Foundations2026-07-25 · AI Governance Working Papers
A working paper series on AI governance and civilizational foundations.
- 哲学的失位与AI伦理的困境2026-06-16 · AI Ethics & Governance
After entering the AI industry, philosophy remains governed by capital and product logic: it has a seat at the table but no power to define. Fragmented ethics may evolve into systemic civilizational risk within highly coupled AI networks.
- Governance Stability under Technological Acceleration2026-05-12 · AI Ethics & Governance
A governance-stability framework under technological acceleration: instead of starting from technical compliance or isolated regulatory mechanisms, it examines the structural conditions that produce governance instability.
- WP–AI–04 AI's Greatest Risk Is Not Singular2026-05-11 · AI Governance Working Papers
AI's greatest risk lies not in a single disaster scenario or isolated technical failure but in the gradual amplification of instability across the interconnected levels of human systems. This is the full-text archive of the LinkedIn public post under the same title.
- WP6 AI Is Not a Technology Problem2026-05-08 · AI Governance Working Papers
The concluding installment of the WP series: the essence of the AI challenge is not a technical problem. The mismatch lies at the formation layer, not at runtime. It summarizes the previous five installments and reframes the nature of the AI challenge.
- WP5 — Structural Limits of AI Systems: From Fixing to Formation2026-05-02 · AI Governance Working Papers
The structural limits of AI systems: "fixing" measures such as alignment, optimization, and control can only improve performance; they cannot change how a system was "formed." The move must be from fixing to the formation layer.
- WP4 — Beyond Systems: The Problem of Formation2026-04-22 · AI Governance Working Papers
The WP4 series: system failure as the expression of inconsistent formation conditions. Technological structures, human factors, and institutional constraints each form according to their own internal logic and are only afterward combined into a unified system, producing "interaction without unity." Optimization at the operational level cannot change the conditions from which outcomes arise.
- WP–AI–CA2: Assessing Canada's AI Sovereign Compute Initiative: Necessity, Limits, and Strategic Opportunity2026-04-21 · AI Ethics & Governance
An assessment of Canada's "sovereign compute infrastructure" initiative: it not only argues for its necessity but analyzes its inherent limitations and strategic opportunities under the centralization of the global AI landscape.
- WP3 — The Limits of Systems Structural Instability in Artificial Intelligence2026-04-18 · AI Governance Working Papers
The WP3 series: the limits of "stability" in AI systems. Stability is not an intrinsic property of systems but an effect of observation: metrics produce local coherence within the observable domain while unmeasured dimensions keep drifting. The more stable the surface, the more the deep structure depends on its adjustment trajectory.
- WP-AI.1 — AI Governance Landscape: A Comparative Analysis2026-04-09 · AI Governance Working Papers
The core framework paper of the parallel WP-AI series: a comparative analysis of the AI governance landscape; an independent series running parallel to the "generative-structure" series.
- WP1 — Rethinking Artificial Intelligence: The Problem of Naming and Conceptual Boundaries2026-04-09 · AI Governance Working Papers
A rethinking of the naming and conceptual-boundary problems of "artificial intelligence": a structural diagnosis of global AI governance.
- WP2.2 — Why AI Ethics Fails: Structural Misalignment2026-04-08 · AI Governance Working Papers
The second installment of the WP2 series: AI ethics fails not for lack of principles but because of structural positioning. Ethical frameworks always intervene after system formation and never reach the structural conditions of system generation.
- WP2.3 — Toward a Generative Grounding of AI Ethics2026-04-08 · AI Governance Working Papers
The third installment of the WP2 series: the two-layer structure of AI ethics (foundational ethics at the generative layer plus dynamic ethics at the development-and-deployment governance layer). A globally unified baseline of foundational ethics is the necessary groundwork of AI governance.
- WP2.1 — AI Ethics Today: Global Frameworks and Their Assumptions2026-04-06 · AI Governance Working Papers
The first installment of the WP2 trilogy: it examines the "position" of ethics within AI systems and governance structures. Comparing frameworks and scholarly approaches across countries reveals a consistent pattern: ethics is mostly introduced after system formation, presupposing system completeness and the like, presenting an overall discrete landscape with no shared starting point.
- AI Governance in Asia: Structural Variants Beyond Western Models2026-04-04 · AI Ethics & Governance
A study of structural variants in Asian AI governance that moves beyond Western models. It is explicitly affiliated with the "AI Governance Structural Analysis" series of the Harmondeg Institute for Philosophy & Practice, founded by the author.
- Re-Grounding AI Governance in the United Kingdom: Strategic Directions to Strengthen the UK's Pro-Innovation AI Strategy2026-04-02 · AI Ethics & Governance
A policy proposal: it examines the structural paradoxes of the UK's current AI governance framework and offers practical strategic recommendations for strengthening its "pro-innovation" AI strategy (including an updated version on structural policy implications).
- Working Paper №1: Global AI Governance — A Structural Diagnosis and Re-Grounding2026-04 · AI Governance Working Papers
A comparison of EU, UK, Canadian, and Asian governance models: existing governance remains at the outcome layer rather than the system formation layer.
- AI Governance in Canada Is Misaligned: Accountability Without Structural Alignment2026-04 · AI Governance Working Papers
Canada's classification-, compliance-, and ex-post-accountability-centered governance cannot keep up with dynamically forming AI risks.
- WP2.1 AI Ethics Today: Global Frameworks and Their Assumptions2026-04 · AI Governance Working Papers
The first installment of the WP2 series: it surveys global AI ethics governance frameworks and their underlying assumptions, noting the lack of structural alignment between ethical principles and AI system formation processes.
- WP3.2 — The Illusion of Stability in AI Systems2026-04 · AI Governance Working Papers
How feedback loops maintain surface stability through adaptation while continuously driving structural drift.
- WP4.1 — System Failure as a Problem of Formation2026-04 · AI Governance Working Papers
Contemporary governance forms technological structures, human factors, and institutional constraints separately and only then combines them, producing persistent tension. System failure is essentially a problem of formation.
- WP4.2 — Why Systems Cannot Produce Coherence2026-04 · AI Governance Working Papers
Systems built from separation can only organize interaction and improve performance; they cannot produce coherence. Coherence must come before the system.
- WP4.4 — From Coherence to System Formation2026-04 · AI Governance Working Papers
System formation does not begin with isolated components but with the conditions that make relationship and continuity possible.
- The Missing Layer in EU AI Governance: Design-Level Risk, Structural Instability, and Policy Recommendations2026-03-28 · AI Governance Working Papers
AI governance frameworks are missing the intermediate link of "design-level risk": risks are generated dynamically along the causal chain from design to system evolution to application. It offers five policy-expansion recommendations.
- AI Risk Beyond Ethics: A Structural Framework for Design and Stability2026-03-27 · AI Ethics & Governance
It argues that AI risk discussion must not stop at the levels of ethics, governance, and alignment: risks arise not only at the application layer but also at the design layer and in system-stability structures. It proposes a structural framework oriented toward design and stability.
- From AI Naming to Governance: Reframing Artificial Intelligence as “Imitative Intelligence System (IIS)” for Civilizational Clarity2026-03-24 · AI Ethics & Governance
It proposes replacing the name "AI" with IIS (Imitative Intelligence System, 智机), to avoid wrongly attributing subjectivity to computational systems and to restore human responsibility.
- From AI Naming to Governance2026-03-24 · AI Ethics & Governance
Global explorations of renaming "artificial intelligence" expose a naming misalignment: "intelligence" implies understanding, consciousness, and responsibility, while AI is only computational architecture plus statistical inference. The article proposes "Imitative Intelligence System (IIS)" as the standard English name and 智机 as the Chinese name, returning governance to a clear and workable foundation of responsibility.
- The Five Structural Paradoxes of Current AI Governance2026-03-17 · AI Ethics & Governance
It identifies five structural paradoxes of current global AI governance: even when governance frameworks are built on rational abstraction, legal formalization, and procedural refinement, they cannot stop the increasingly irreversible structural instability.
- The EU AI Act and the Emerging Governance Generation Gap2026-03-06 · AI Governance Working Papers
Using the EU AI Act as a case study: a "governance generation gap" is forming between static governance structures and rapidly evolving AI systems.
- AI Governance as a Civilizational Positioning Problem2026-02-23 · AI Ethics & Governance
Using civilizational positioning as its analytical framework: it places AI governance on the horizon of the center of responsibility, examining the structural relationship between technological amplification and human accountability as systemic impact expands.
- When Consequences Sink: Responsibility Misalignment and the Illusion of Stability in Modern Systems2026-02-09 · AI Ethics & Governance
It introduces the concepts of "consequence sinking," structural absorption, and decision-outcome decoupling: modern systems appear stable yet lose their capacity for self-correction through the misalignment of responsibility.
- The most dangerous risks to civilization look highly rational2026-02-08 · AI Ethics & Governance
If the necessary negation and trial-and-error methodology of the natural sciences is transplanted into the domain of ethical and civilizational judgment, it becomes a hidden civilizational risk that dissolves judgment and responsibility. Natural principles and social principles must be strictly distinguished in practice.
- Artificial Intelligence and Human Civilization2026-02-03 · AI Ethics & Governance
It proposes a civilizational framework for AI governance and human responsibility, with an appended Harmondeg core glossary (Harmondeg defined as a civilization-oriented framework that maintains civilization's cross-temporal coherence, sustainability, and non-destructive continuity).
- AI Governance Is Not a Technical Problem – It's a Civilizational One2026-02 · AI Ethics & Governance
Governance failure stems from capability expansion outpacing ethical and responsibility structures. It proposes positioning, responsibility anchoring, asymmetric correction, and reversibility protection.
- NEW BOOK: AI Governance Must Move Upstream—to the Formative Layer· AI Governance Working Papers
Announcing the Amazon release of the new book The Governance Impasse of Artificial Intelligence (paperback and Kindle), published by the Harmondeg Institute for Philosophy & Practice: it argues that AI governance must move upstream to the Formative Layer, providing earlier signals, judgment, and opportunities for human intervention before instability fully manifests.