Mohammed on AI Legal Personality and the Corporate Analogy

Dadvan Mohammed (Universiti Utara Malaysia) has posted Rethinking AI Legal Personality: From Human Analogy to Corporate Analogy on SSRN.  Here is the abstract:

The rapid proliferation of autonomous artificial intelligence systems across commercial, governmental, and social domains has exposed a structural crisis in legal attribution — the “responsibility gap” identified by Matthias — whereby existing frameworks fail to connect AI-generated harm to identifiable responsible actors. This article argues that the dominant scholarly response to this challenge, the human analogy, is theoretically deficient and practically unworkable. Grounding AI legal personality in philosophical criteria of consciousness, volition, or moral agency imports requirements that the legal system has never applied to non-natural persons and has no institutional competence to assess reliably.

This article proposes a reorientation of the jurisprudential debate from the human analogy to the corporate analogy, drawing on functional legal theory as developed by Kelsen, Dewey, and Hart. Legal personality, properly understood, is an instrumental regulatory construct conferred by the law in service of identifiable social and economic purposes. The commercial company — as established in Salomon v A Salomon & Co Ltd [1897] AC 22 — provides the definitive historical precedent: a non-conscious, non-volitional entity accorded full legal personality on purely functional grounds.

Building on this foundation, the article proposes the Functionally Limited AI Legal Personality (FLAILP) model, structured around five core characteristics: functional limitation, conditionality on operational autonomy, independent financial liability, mandatory governance and oversight, and a novel doctrine of Piercing the AI Veil. The model is applied across tort law, contract law, and regulatory accountability, and situated within the EU AI Act 2024, the OECD Principles on AI, and emerging national frameworks. Critical objections — including the consciousness objection, liability evasion risk, and the limits of the corporate analogy — are addressed with rigour. The article concludes that FLAILP represents not a departure from legal tradition but its most faithful contemporary application.

To receive new posts from Legal Theory Blog by email, get a free subscription to Legal Theory Stack.

Lawrence Solum