Intellectual Property Law in the Age of Generative AI: Challenges in Patentability and Copyright
Abstract
The relationship between artificial intelligence and intellectual property law has entered a critical phase of
evolution. Generative AI—defined as machine-learning systems capable of autonomously creating text, art,
music, code, and inventions—has raised fundamental questions about authorship, inventorship, ownership,
and originality. For over a century, intellectual property systems across the world have operated on the
premise that creativity, ingenuity, and invention originate from human intellect. The twenty-first-century
revolution of generative algorithms such as OpenAI’s GPT architecture, DeepMind’s AlphaCode, Stability
AI’s Stable Diffusion, and Google’s MusicLM challenges that anthropocentric assumption. These models
produce novel outputs without direct human authorship, threatening to destabilize the conceptual
foundations of copyright and patent law.
This research examines how generative AI’s capacity for autonomous creativity tests the boundaries of
existing intellectual property regimes. It explores the central legal question: can a non-human entity be
recognized as an author or inventor, and if not, how should the law allocate ownership of AI-generated
works? The inquiry unfolds across two interrelated domains: patentability, which depends on novelty,
inventive step, and industrial applicability, and copyright, which hinges on originality, authorship, and
fixation. The objective is to evaluate whether current legal standards under international treaties—such as
the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS, 1994), the Berne
Convention (1886), and the Paris Convention (1883)—and national legislations such as India’s Patents Act,
1970, and Copyright Act, 1957, are equipped to handle algorithmic creativity.
evolution. Generative AI—defined as machine-learning systems capable of autonomously creating text, art,
music, code, and inventions—has raised fundamental questions about authorship, inventorship, ownership,
and originality. For over a century, intellectual property systems across the world have operated on the
premise that creativity, ingenuity, and invention originate from human intellect. The twenty-first-century
revolution of generative algorithms such as OpenAI’s GPT architecture, DeepMind’s AlphaCode, Stability
AI’s Stable Diffusion, and Google’s MusicLM challenges that anthropocentric assumption. These models
produce novel outputs without direct human authorship, threatening to destabilize the conceptual
foundations of copyright and patent law.
This research examines how generative AI’s capacity for autonomous creativity tests the boundaries of
existing intellectual property regimes. It explores the central legal question: can a non-human entity be
recognized as an author or inventor, and if not, how should the law allocate ownership of AI-generated
works? The inquiry unfolds across two interrelated domains: patentability, which depends on novelty,
inventive step, and industrial applicability, and copyright, which hinges on originality, authorship, and
fixation. The objective is to evaluate whether current legal standards under international treaties—such as
the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS, 1994), the Berne
Convention (1886), and the Paris Convention (1883)—and national legislations such as India’s Patents Act,
1970, and Copyright Act, 1957, are equipped to handle algorithmic creativity.