Ownership of AI-Generated Works: Who owns content created by AI? The User, Developer, or No One? Examine gaps in Copyright Law

Abstract

Text, image, music, and software code creation systems that can autonomously produce content have revealed a serious loophole in modern copyright law. The main issue this paper will discuss is how the authorship and ownership of AI-generated creative works can be established, which is still lacking, causing a great deal of legal ambiguity to developers, users, courts, and policymakers.

The article is based on a doctrinal research approach to law, which involves a synthesis of comparative statutory law practice, case law, and an assessment of new policy tools in the United States, the United Kingdom, the European Union, and China.

Three underlying tensions are found. To begin with, the doctrine of human authorship simply denies AI-generated works the protection of copyright, leaving their ownership void. Second, conflicting interests between AI creators, end users, and advocates of the public domain are impossible to reconcile and require a solution at the legislative level instead of the purely judicial level. Third, cross-jurisdictional lack of consistency - exemplified by the difference between U.S. and UK practices - creates piecemeal protections that cannot suit the global scope of AI-produced material.

The article suggests a tiered ownership structure that is adjusted to the level of human creative input, augmented by a sui generis tool of developer interests and a compulsory disclosure duty, and provides a principled way of achieving legal certainty.