Reforming Pakistan’s Copyright Law for Generative AI: A Human Creative Control Framework
Abstract
Pakistan's Copyright Ordinance 1962 does not expressly address AI-assisted works, autonomous machine output, prompt-based creation, training data or platform responsibility. This paper develops a reform framework suited to Pakistan's legal system and digital economy. It uses doctrinal and normative analysis informed by comparative approaches to identify gaps in authorship, originality, ownership, registration, evidence, training-data licensing, liability and remedies. The paper recommends that AI systems should not be recognized as authors; copyright should protect only identifiable human expression in AI-assisted and hybrid works; and substantially autonomous output should remain outside ordinary copyright. It proposes a Human Creative Control Test focused on human choices, the role of AI as a tool, intellectual shaping of the final work, identifiable contribution and non-infringement. Reform should also require proportionate disclosure of material AI use, preserve prompt and editing evidence, support collective licensing for commercial training, create limited research exceptions based on lawful access, and allocate liability according to control, knowledge and benefit. A staged implementation led by IPO Pakistan, followed by statutory amendment and institutional capacity building, can protect creators while supporting responsible innovation.
How to Cite
Monika, Bhoma Ram Ji, "Reforming Pakistan’s Copyright Law for Generative AI: A Human Creative Control Framework", Vol. 4, Issue 1, 24-04-2026, pp. 116-126. DOI: 10.5281/zenodo.21350837
