THE ALGORITHMIC INVENTOR AND THE ANTIQUATED STATUTE: RE-ENGINEERING THE INDIAN PATENT SYSTEM IN RESPONSE TO AUTONOMOUSLY GENERATED AI INVENTIONS

AUTHOR – RIDDHI JAIN, BBA LL.B. (H.), AMITY LAW SCHOOL, NOIDA

BEST CITATION – RIDDHI JAIN, THE ALGORITHMIC INVENTOR AND THE ANTIQUATED STATUTE: RE-ENGINEERING THE INDIAN PATENT SYSTEM IN RESPONSE TO AUTONOMOUSLY GENERATED AI INVENTIONS, INDIAN JOURNAL OF LEGAL REVIEW (IJLR), 6 (8) OF 2026, PG. 576-587, APIS – 3920 – 0001 & ISSN – 2583-2344. DOI – https://doi.org/10.65393/IJLRV6I860

ABSTRACT

The emergence of Artificial Intelligence (AI) as a capable creator of novel and patentable inventions has exposed significant limitations within the Indian patent framework. The Patents Act, 197o, is fundamentally premised on human inventorship, thereby rendering it inadequate in addressing inventions autonomously generated by AI systems. This paper critically examines the concept of the “algorithmic inventor” and highlights the legal vacuum concerning inventorship, ownership, and accountability in such cases.¹

Historically, intellectual property regimes were constructed around the “Romantic Author” or “Human Inventor” theory—the belief that innovation is a byproduct of human cognitive labor and subjective “sparks of genius.” However, the transition from AI as a mere tool to AI as a generative agent has fractured this human-centric paradigm. This research explores the doctrinal tension between the “Mental Act” requirement in patent law and the computational reality of deep learning, where inventive outputs are derived through non-linear data processing rather than human biological thought.

The analysis further extends to international developments, particularly the landmark “DABUS” (Device for the Autonomous Bootstrapping of Unified Sentience) cases across several jurisdictions. These cases serve as a global litmus test, revealing a profound hesitation in recognizing non-human entities as inventors due to statutory constraints and public policy concerns regarding liability and personhood. In the Indian context, the paper scrutinizes the interplay between Section 3(k) of the Patents Act, which bars algorithms per se, and the evolving standards of Computer-Related Inventions (CRIs).

By identifying the “Black Box” challenge—where the internal logic of an AI system remains opaque to the human eye—the paper underscores the growing difficulty in satisfying the statutory requirements of “Enablement” and “Disclosure.” The research argues for a strategic “re-engineering” of the patent system. It proposes a shift toward a hybrid inventorship model, the introduction of sui generis protections for machine-generated works, and a recalibration of the “Person Skilled in the Art” (PHoSITA) standard. Ultimately, this study advocates for a flexible, forward-looking legal architecture that balances the necessity of incentivizing AI-driven technological leaps with the essential need to preserve human accountability and the public interest.²