Skip to main content

Can computers and AI systems really be inventors?

A law professor at the University of Surrey is arguing that it should be possible for computer-based artificial intelligence (AI) systems to be formally considered as inventors for any invention they contribute to, much in the same way a person would. The argument forms part of a paper, which has been published in the Boston College Law Review, entitled I Think, Therefore I Invent: Creative Computers and the Future of Patent Law. 

In its introduction the report makes the point that while inventions by computers have been granted patents previously, the concept of computer inventorship has never actually been considered by the courts. The concept of giving creative computers the credit for their own inventions may sound surreal but, in reality, they have been generating potentially patentable ideas for decades without acknowledgment.

The issue of ownership

As Professor Ryan Abbott points out in his paper, ‘machines have been autonomously generating patentable results for at least twenty years and the pace of such invention is likely increasing.’ The fact that such innovation is proliferating makes it even more important that the issue of ‘ownership’ is resolved, or uncertainty could prevent it from being published in the future and potentially give rise to disputes. One well known example of creative computing in action is ‘The Creative Machine’, a computational paradigm invented by computer scientist, Stephen Thaler in 1994. 

This machine is capable of creating novel patterns of information and adapting to different scenarios without human intervention. Among its outputs, the machine has been credited with inventing a prominent design aspect of the Oral-B CrossAction toothbrush. Other examples of creative computing include a software-based system modelled on the process of biological evolution known as genetic programming and IBM’s ‘Watson’. Described by IBM as a machine capable of generating ‘ideas the world has never imagined before’, Watson utilises algorithms to generate novel ideas based on pre-determined parameters. In its latest incarnation, Watson is able to generate potentially-patentable recipes based on user preferences and the predicted quality of the dish. Recognising that AI systems are indeed capable of inventing things is an important first step to finding a solution to the problem of ownership. Until now, it has been advised that a person who 'discovers' a solution invented by such a system should present themselves as the named inventor of the solution in a patent application. But is this the right thing to do, or could it have an adverse effect for inventions that stem from this increasingly important area of research science in the future?

Lack of clarity

Under current patent law, patent applications must include one or more named inventors who must each be a 'person' or ‘individual’, rather than a corporate entity. To qualify as an ‘inventor’, an individual must have actually devised part of the inventive concept. It seems that AI systems such as The Creativity Machine are clearly capable of operating in this way, without human intervention. This uncertainty over who should be considered an inventor of an AI-generated invention raises some interesting questions over ownership of a patent for the invention. 

Human inventors would be entitled to own the patent unless they have assigned the right to a new owner or are employed in a position which requires them to assign the invention to the employer. However, if an AI system is capable of being named as an inventor, who has the right to an invention it devises?  From a patent law perspective, it is easy to see how the current lack of clarity regarding the ownership of artificially-generated inventions could lead to disputes. 

Take the scenario where Person A creates an AI system, which is capable of autonomously generating novel ideas, and licenses it for use to Company B. The AI system outputs a blueprint for a new device, which is first seen by Person C in Company B's office and person C realises that the new device is ground-breaking. In such a scenario, who is the inventor: the AI system or Person C? Furthermore, who is the owner of the invention; the creator of the AI system, Person A, Company B or Person C? 

Would the licence between Person A and Company B have to cover both eventualities to guarantee a desired chain of ownership from Person A to Company B, or retention with Person A? If Person C is deemed to be the inventor and is employed by Company B in a position from which no invention is expected, Person C could claim to be the owner of the invention. If Person C is a non-contracted intern, then the act of seeing the blueprint could be considered a public disclosure that would prejudice a patent being filed.    

Huge economic implications

If there is continued uncertainty over ideas generated by AI systems, it is possible that activity in this area could be stifled as a result. After all, the inability to own and validly patent such inventions would in many cases make it virtually impossible to leverage them commercially by attracting private sector funding for their further development or licensing the rights to a third party. As the report points out, without any clear right of ownership, the ideas generated by AI systems and computers may not be disclosed at all, for fear that others might use them. 

Instead of inventions being shared with the world via the patent system, they might be withheld as trade secrets until there is a greater degree of clarity. For society to benefit from the rich seam of innovation activity now taking place in the field of computer science, the issues surrounding rights of ownership must be resolved sooner rather than later. For this reason, Professor Abbott’s report is well timed and raises an important issue, which could have economic implications for us all. 

At the moment, the thought of crediting a computer or AI system as an ‘inventor’ could seem fanciful (to the uninitiated), but a potentially game-changing invention could be just around the corner. It would be a shame for the momentum generated by such an important innovation to be lost in a battle over IP rights.    

Stuart Latham and Diego Black,  Withers & Rogers LLP.
Image Credit: PHOTOCREO Michal Bednarek / Shutterstock

Diego Black is a senior associate at intellectual property firm, Withers & Rogers LLP. He specialises in advising businesses in the field of consumer electronics and computing.