AI made Inventions, who owns these?
With the advent of artificial intelligence (AI), the end of patent law is near. Though it may not happen today or tomorrow, the system’s decline is underway. Groundbreaking innovations in AI technology have made inventions “made by AI” a reality. Today, AI is able to “invent” not only new materials and machines but also manufacturing processes, pharmaceutical drugs, and household products. Soon, our life will be replete with artificial artifacts. In a sense, humans no longer stand at the center of the creative universe — we are no longer the masters of innovation.
Despite this upheaval on the horizon, the consequences for legal doctrine remain largely unexplored. The most fundamental disruption we are about to witness is the emergence of scenarios where the inventive input is made by AI — and where no human “fire of genius” is involved. The problem here is that patent doctrine is still founded on the idea that only human beings can be inventors. We desperately need a patent update for the era of artificial inventions.
With regard to scenarios of autonomous AI inventions, it is puzzling that terminology is not uniform. The phenomenon is still mostly described as “computer-aided,” “computer-based,” or “computer-automated” inventing and its results as “computer-generated inventions.”. Some academics call the outcome of autonomous invention by a computer or an AI application a “computational invention.
Stage One: At the very left (I human), we may picture the epitome of the “classic” inventor. Here, any inventive output is genuinely human-made. The creativity remains an intrinsic and genuinely human feature of the inventive system that consists of the human inventor and their tools. Consequently, the inventor’s capacities with respect to the actual solution will not increase so significantly that patent law would have to adapt the standard for “non-obviousness.”
Stage Two: The situation changes once the inventor is supported by more elaborate technology. The era of technology-supported or technology-assisted invention started with the use of early computing machines, mainly for collecting and processing data. One might consider this the very first hybrid system of human-AI invention. The inventions in these scenarios are still considered genuinely and exclusively human-made.
Stage Three: It is the point at the demarcation line between the fields of partially AI-supported and autonomous AI invention scenarios where the most serious doctrinal distortion looms (between I human-AI and I AI). As can be seen, in the area of I human-AI, there will be a point where the inventive input necessary to overcome the threshold of non-obviousness will no longer consist of contributions by both the human actor and the AI application. At this point, AI application has taken the driver’s seat with regard to conceiving of a solution. Therefore, the inventive input necessary to overcome the non-obviousness threshold will be provided by the AI application alone. As a look at the technological foundations has shown, many practical scenarios exist in which the human actor’s role in an inventive process is reduced to a mere onlooker.
Finally, at the very right of the area of autonomous AI invention (I AI), we will find the ultimate step in which an ultra-intelligent AI has not only taken over the position as actor conceiving of the invention but also become a supervisor of its own volition. In these scenarios, AI will ask the questions as well as provide the answers. The role of human actors — if they continue to play an active role at all — remains to be seen.
Final Thoughts
The World Intellectual Property Organization recently questioned whether it might be too early for a definite change in the legal system, since the impact of AI is still unfolding at a rapid rate and our understanding of the effects of different policy measures may still be insufficient.
We will see that the existence of such “inventions without an inventor” heralds the need for changes to patent doctrine. The issue will be: If there is no human inventor, who should qualify? Accordingly, it is disputed as to who best deserves to own rights to emergent inventions or where best to allocate such rights under an economic perspective. Among the candidates for right ownership are the AI’s developers or programmers, its owners, and its users.
Humankind is not Mother Nature’s last word on the question of intelligent life. To the contrary, much is likely yet to come. Why, therefore, should we worry about the fact that our current intellectual capacities are just a snapshot in history? Isn’t it evident that homo sapiens will one day be outdone by superior minds? And who ever said that these minds cannot be artificial?
Reference: Dornis, T.W., 2020. Artificial Intelligence and Innovation: The End of Patent Law As We Know It. Yale Journal of Law & Technology, 23.
