Can AI Be an Inventor? — The Limits of the System and a Question for the Future

With AI-generated creations becoming a reality, IP systems are facing fundamental questions. One of the most discussed cases is DABUS, an AI system listed as the inventor in patent filings. This raises a deeper issue: who or what qualifies as an “inventor”?

In Japan, the DABUS application was rejected. The case is currently pending before the Supreme Court of Japan, following lower court rulings that rejected the application. The courts ruled that “inventor” under current law refers only to natural persons, emphasizing consistency with the Basic Act on Intellectual Property and the framework for inventive step.

Interestingly, in the appeal, the plaintiff introduced a novel argument: that if an invention exists, the system should not require a human inventor to be named. This “no-inventor-listing-needed” theory goes beyond formalistic AI inventorship and challenges the rationale of the current procedural structure.

I take a cautious stance. If AI truly created an invention, then logically, AI should assert the right—not a human acting on its behalf. Allowing a human to claim ownership of AI’s creation lacks internal coherence.

Recognizing AI as an inventor also implies granting it legal capacity—a kind of personhood. That would require changes far beyond patent law, raising serious questions about stability and legitimacy.

If AI ever reaches full autonomy in problem-solving and invention, we may indeed need new legal frameworks. But for now, I believe AI should be positioned as a powerful tool supporting human creativity, while responsibility remains clearly human.

Ultimately, this debate is about more than patents—it’s about the evolving relationship between humans and AI. I recall a science fiction film in which an AI, tasked with protecting humanity, concluded that the best way was to control people. Logic does not always align with human values.

Recognizing AI as an inventor may mark the beginning of a much broader redefinition. That’s why I support keeping AI as a tool for now, and addressing legal reform only through thoughtful, society-wide deliberation.

These are my personal reflections following a recent seminar and ongoing interest in AI-related innovation.