Inventors build things.
They solve problems, file patents, and crucially get recognised for it.

But what happens when the “inventor” isn’t human?

That’s exactly what courts around the world have been grappling with, thanks to one very controversial AI system: DABUS.

In Thaler v Comptroller-General of Patents Designs and Trade Marks, Dr Stephen Thaler tried to list his AI (DABUS) as the inventor of two patent applications.

His argument was simple:
if AI created the invention, it should be named as the inventor.

The court disagreed.

The ruling confirmed that under current UK law, an inventor must be a natural person. No exceptions.

Not Just a UK Problem

This wasn’t a one-off.

Similar applications were filed and rejected in multiple jurisdictions, including the US and Europe. Courts have been remarkably consistent:

• AI cannot be recognised as an inventor
• Only humans can hold that legal status
• Ownership flows from human inventorship, not machine output

So, for now the law is clear. But the debate is far from over.

Here’s where things get tricky for patent law:

• If a human simply presses a button, are they really the inventor?
• If AI generates a novel solution no human anticipated, who gets credit?
• If no human contribution is meaningful enough, does the invention fall apart legally?

Patent law is built on the idea of human ingenuity.
AI is starting to blur that line.

Why the Law Is Holding the Line

There’s a reason courts are reluctant to recognise AI as inventors:

• Legal systems rely on accountability, AI can’t own rights or bear responsibility
• Patent frameworks are designed to reward human effort
• Expanding inventorship to AI could disrupt ownership, licensing, and enforcement

In short, recognising AI as an inventor creates more problems than it solves—at least for now.

This debate isn’t going away. If anything, it’s getting sharper:

• Will laws evolve to recognise AI-assisted inventorship instead?
• Could we see a new category; somewhere between human and machine?
• How will courts assess “sufficient human contribution” in the future?

Right now, the answer is simple:
AI can help invent but it can’t be the inventor.

But as AI systems become more autonomous, that line will get harder to defend.

And when it does, patent law won’t just need interpretation.
It’ll need reinvention.