The intersection of AI and IP rights: Where do we draw the human line?

For the last three years, intellectual property (IP) offices worldwide have been overwhelmed with “non-human” applicants, such as autonomous algorithms, large language models (LLMs), and diffusion networks that lay claim to the titles of “author” and “inventor,” terms that were previously hallmarks of human achievement. In light of this, a new global consensus has hardened into what jurists describe as the “Human-Only” standard.

Across the key hubs of innovation—the Americas, Europe, Japan, and India—the verdict is clear: while AI can be a powerful brush, a tireless lab assistant, or a sophisticated calculator, it currently cannot own the outputs it generates. As of early 2026, the legal “soul” of innovation remains strictly biological.

The legal limits of algorithms

In this context, the principal challenge has come from Dr Stephen Thaler and his AI system, DABUS (Device for the Autonomous Bootstrapping of Unified Sentience). In the United States, the landmark judgment in Thaler v. Vidal, established that the Patent Act’s use of the word “individual” refers exclusively to natural persons. The US Supreme Court’s refusal to grant certiorari in 2024 effectively brought this debate to a close.

On January 30, 2025, the Intellectual Property High Court of Japan reaffirmed that a creation must be the product of “human creative activity” for it to qualify for patent registration. The court also noted that since AI lacks the legal capacity to hold rights or bear liability, it cannot be recognised as a patentee.

India, following suit, has been no different in its approach. The Delhi High Court upheld the Controller General of Patents’ rejection of DABUS, citing Section 6 of the Patents Act, 1970. The section requires an applicant to be a “person”, not an algorithm.

If inventions reflect the survival instincts of humankind, literary and artistic indulgences are often said to reflect the soul. Here, as well, the law has drawn a clear line.

Under the “Human Touch” doctrine, the US Copyright Office (USCO) reinforced its stance in late 2025. In cases involving AI-assisted works, most notably Allen v. USCO, regarding the award-winning Théâtre D’opéra Spatial, the Office ruled that even highly specific prompt engineering does not amount to “authorship” when the AI independently determines the expressive elements of the output.

In conclusion, these developments point to the fact that the law has drawn a circle around the human, recognising that “intelligence” is a commodity, but “creativity” is a legal privilege. The “Human-Only” standard isn't just about protecting jobs; it’s a philosophical standard. By denying AI the right to own, the law ensures that control remains with humans.

Reported by Khushraj Singh, Senior Reporter, The Global IP Magazine. Email Khushraj: newsdesk@northonsprmarketing.com