OWNERSHIP OF THE ALGORITHMIC VOICE: ALLOCATING ECONOMIC RIGHTS IN AUTONOMOUS AI-GENERATED LITERARY WORKS UNDER PAKISTANI COPYRIGHT LAW

Abstract

Pakistan's Copyright Ordinance 1962 does not contain a rule that identifies the author of a genuinely autonomous AI-generated literary work or allocates first ownership where no legally cognisable human author can be identified. The problem begins with section 2(d), which defines the author of a literary work as 'the author of the work', and continues in section 13, whose first-ownership rule and employment exceptions presuppose an author. This article compares that position with the human-authorship approach in the United States, the statutory computer-generated-works rule in section 9(3) of the United Kingdom's Copyright, Designs and Patents Act 1988, and the 'person who causes the work to be created' formulation in section 2(d)(vi) of India's Copyright Act 1957. It retains the proposed Company Model but places it on a narrower legislative foundation. The model does not treat an AI system as an author and does not convert corporate investment into creativity. For the residual category in which sufficient human authorship is absent, the article recommends a limited statutory economic right, distinct from ordinary authorial copyright, with the qualifying corporate deployer as the default rights-holder subject to contractual reallocation. The deployer is preferred because it can be identified through operational control, market-facing responsibility and contractual accountability, while the rule remains narrow enough to avoid transferring rights in the underlying model or displacing human authorship. Section 13 is used only as a structural analogy for legislative design. The resulting proposal keeps authorship, originality, eligibility for protection and allocation of economic rights analytically distinct.