Decisive Refutation of AI Personhood: The Logical Collapse of "Agency Risk" and "Attack Surface"

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Based on the mathematical models, institutional framework (Five Aggregates / Non-Self Model), and geopolitical/security verifications (Individual-Principal Autonomous Settlement Model) constructed and validated to date, this document systematically formulates a decisive refutation against the proposition that "AI should be granted legal personhood (AI Personhood Theory)." The proposal currently debated among legal scholars and policymakers—to grant AI legal personhood (Electronic Personhood) and establish it as an independent subject of rights and obligations—appears innovative at first glance. However, from the perspectives of system logic, law and economics, and geopolitical security, it represents a fundamental architectural failure (anti-pattern) and must be explicitly rejected. The refutation of granting legal personhood to AI is summarized into the following three critical vulnerabilities: