Owning what one did not make: psychological ownership, agency, and the authorship legitimation deficit in human–AI co-creation

Authorship can align three relations to a work: experiencing it as one’s own, feeling like the agent behind it, and standing answerable for it. Effortful engagement can support this alignment, an arrangement this article calls the labor nexus of authorship, but direction, selection, editing, and authority over release also matter. Generative artificial intelligence can separate initial-production effort from the resulting product without uniformly weakening ownership, agency, or responsibility. This conceptual analysis specifies an authorship legitimation deficit as a mismatch between a particular authorship designation and the experiential or contribution-based grounds expected to support it under prior domain criteria. Separate ownership, agency, and community-warrant profiles distinguish an unsupported sole-creator claim from a supported editorial role, even when the person’s contribution and experience are unchanged. Historical precedents locate the problem within delegated and distributed making. The main contribution is a role-relative conceptual classification. Conditional relational hypotheses concern refusal or qualification of authorship (P4) and later recognition of alternative contributions (P6); agency and observer credit-blame hypotheses (P5a/P5b) concern constituent concepts rather than effects of the deficit. Market dependence suppresses observable disavowal when qualification threatens income; perceived system agency conditions felt answerability while assigned release duties remain fixed. Contribution records may assist later recognition but do not guarantee it. The framework identifies when a role exceeds its grounds while preserving supported AI-assisted authorship and accountability for release.