It is an era when generative AI creates drug candidate substances, autonomously optimizes complex algorithms, and presents combinations far exceeding the limits of human researchers. Now the statement "AI invented it" is becoming not a metaphor but a literal truth.
In this situation, the U.S. Patent and Trademark Office (USPTO) recently drew a firm line through new guidelines.
"AI cannot be an inventor. Invention begins in a human mind."
At a moment when AI's creative capabilities are expanding at the fastest rate in history, the United States has actually fixed the subject of invention as human. This decision is not simply a legal interpretation but a point that makes us deeply consider what humans should be evaluated by in the AI era.
The Moment Technology Replaces Humans, Institutions Retreat to Humanity
AI is already exploring ideas and combining solutions faster than humans in countless research settings. However, the USPTO stipulated that "a human must exist at the final threshold of creation."
The reason is that what the patent system aims to protect is not the result but the human thought process and creative judgment (unique human contribution) that created that result.
In other words, no matter how much technology surpasses humans, the institution will use human unique thinking ability as its standard.
Ironically, the more outstanding AI is, the more important human 'humanness' becomes as an evaluation criterion.
Conception: Humanness as the Last Boundary of Patents
The standard re-emphasized in the USPTO's guidelines this time is Conception. Invention must be a technical solution specifically formed in complete form in a human's mind, not just a simple idea.
However, there are many cases where AI-produced results are so complex and precise that humans 'do not fully understand' their principles. Then what happens to that invention?
The USPTO's answer is cold. Inventions that humans do not understand cannot be patented. Simply reviewing or selecting AI-derived solutions is not invention. Directing AI to "solve this problem" is not a creative contribution.
That is, no matter how excellent an answer AI produces, if there is no human thinking capable of understanding that answer, giving it meaning, and taking responsibility for it, it will not be recognized as invention. At this point, Conception becomes the standard of humanness.
New Challenges Thrown by the Impossibility of 'AI Co-Invention'
The USPTO officially did not recognize co-inventorship between AI and humans. The reason is that since AI is not a 'person' in the legal sense, the co-invention relationship itself cannot be established.
This throws important implications to those of us living and struggling through the AI era.
① AI-led inventions become difficult to protect
If AI created the core solution, if human Conception is not proven, the patent itself may be invalidated.
② The capability to document human creative contribution becomes important
Going forward, companies must reorganize research records centered on humans. Documentation proving 'human creative thinking' — including prompt design, reasons for judgment, modification processes — becomes essential.
③ Discrepancies in global patent strategies arise
Some countries may broadly recognize AI inventions, but the United States is resolute. The moment AI is listed as inventor in an international application, there is also risk of losing U.S. patent priority.
When AI creates new innovations, law is actually demanding human roles more precisely.
The More Powerful Technology Becomes, the Deeper Creativity Required of Humans
In an era when AI 'executes' inventions, the elements needed for humans to 'declare' inventions become more precise.
① Ability to define problems: Human directionality that AI cannot
Problem setting is not a simple command but a process reflecting human experience, values, and purpose.
② Ability to give meaning to results: Interpretation as uniquely human creativity
The answer AI creates itself is not creation. The interpretive work of grasping the meaning of that answer is precisely the domain of human creativity.
③ Judgment of direction, ethics, and responsibility: Decisions only humans can make
Invention is not merely a technical discovery but also an act of judgment considering impact on society. AI has no will and cannot take responsibility. Ultimately, the more AI expands invention, the creativity required of humans is moving to a deeper and more essential domain.
The Patent System of the AI Era Demands Humans to 'Be More Human'
The era when AI invents has already begun. However, the USPTO's guidelines this time deliver a clear message. The subject of invention is not technology but humans. This is not the stubbornness of human-centrism but a confirmation that the values of thinking, judgment, responsibility, and interpretation that only humans can do are still absolute.
As technology surpasses humans, what we should do is not compete with AI but further strengthen the humanness that AI cannot replace — specifically creativity, meaning interpretation, accountability, and directionality.
Even if AI invents, patents belong to humans. The reason is simple. The final judge who assigns 'value' to the answers AI creates is still human.
For that reason, for us humans, the true competitiveness of the AI era can only be humanness rather than technology.
