Granting IPR to A.I.: The Anthropomorphic Conundrum



[This article is co-authored by Prakarsh K & Shruti Mishra, National Law Institute University, Bhopal]

Introduction
Innovation has always been the focal point of human development. It is safe to say that an innovative mind has been more or less reserved for humankind. However, is there a way that we have found to share this gift of nature with an inanimate entity? And if so then what does that entail? 
Artificial Intelligence (Hereinafter referred to as ‘AI’) is the closest to a fabricated consciousness that has been created. Innovation by learning is one of the prime points of research in the field of AI. When we look at innovation from a legal perspective, the first thing that comes into consideration is the rights to protect it, specifically intellectual property rights (Hereinafter as ‘IPR’).
IPR includes various rights and protections given to individuals such as rights to recreate, sell, distribute, exclusive use, etc. It essentially protects the ‘intellect’ that flows from any individual. However, when these were brought into force it was with the consideration that it may only be applicable to a ‘person’. At that point in time, one could not fathom the possibility of what the future holds when it comes to inventiveness. AI has been hailed as the closest link between software and the human mind. It has a myriad of applications and quality, one of them being the ability to learn and evolve according to the environment it is provided. This has a humanizing effect as it can based on its learning go beyond what was coded into it to create something new. The creation can be unique, novel and will have its practical applications; considering these it may very well qualify to be registered as intellectual property.  
AI systems can be registered under their individual creators, however, what about the creation that comes from an AI? Who will it be registered under and who will have the rights to it is the moot point. 

The Humanizing Factor
One of the requirements for requesting a patent is for it to be done by a ‘person’ according to the Patents Act, 1970. [1] A ‘person’ may not necessarily be a biological term (Except for the U.S. where a patent can only be requested by a natural person.), [2] it can mean any entity having the status of a legal person which may include companies, firms, etc. [3] This establishes that it is not necessary to be made of blood and bone to be considered as eligible to hold the license and rights to an IPR.
Legal personality being granted to AI has already started being considered in various countries. [4] This consideration did not initially stem from the possibility of granting AI the rights for creations, instead, it was first discussed due to the criminal activities that AI might indulge itself into. To pin the liability on the doer or the enabler or creator was the question. While that was being discussed on the criminal front the civil branch started out in a positive direction discussing the possibility of giving AI rights. It was at this point that the possibilities in IPR were shed light on. 
It is pertinent to note here that it is not being suggested that AI be given the same status as a human being or anything similar, that would involve issues of moral and ethical character which is a totally different area. It is more of a corporate version of legal personhoods such as one granted to firms or organizations giving them limited rights and imposing liabilities.
It may also be a possibility that a new category may be created for AI as being a non-living entity but having a certain level of autonomy. Without the considerations of human sentiment but of human or even higher intellect. These may be included in a Statue bringing it under existing legislation or may be subject to an entirely new policy, either way, that is in the distant future. 

No Strings Attached: AI Autonomy  
It is no surprise that currently there exist no provisions whatsoever to grant AI intellectual property rights. 
“Without human intervention and training, AI algorithms will churn out the rubbish.”
Currently, the law does not see AI as an independent entity, it sees it as an intermediary between a human and the desired result. An aiding mechanism that makes data processing and such activities faster and easier. However, this is no more completely true and definitely will not be in the near future. 
Instances of autonomous AI innovations and actions are increasing with each passing day. There are instances wherein a commercially available AI system can integrate AI and machine learning dimensions to almost any software programme and process large amounts of data without any organized input. [5] All this without any actual human intervention. The initial product was a human creation, however, what about the software code it evolves on its own into something new. Further in the field of medicine and biotechnology scientists have been using AI to expedite the process of arriving at medicines and vaccines. Currently, this requires structured input of data however with the level of development this requirement has been decreasing. [6] AI is able to study biological data and arrive at inputs without any intervention.
AI is also used by reporting agencies to create pieces of work that they publish by simply allowing the AI to scourge through available material which is the exact same thing any other employee would have done prior to availing AI services. [7] 

Worthy Enough? IPR & AI
There have been cases across the globe wherein it was discussed that there is the requirement of human uniqueness when registering for IPR. [8] However, over the years these vague terms were overruled by judgments that watered down the requirements to be limited to the mere fact that it is not copied. In India, even derivative work may be considered as intellectual property as long as there can be a distinction that is drawn. [9] Derivative work is something AI has been used for time and again mainly used by news reporting agencies. [10] This presents a rather significant possibility of AI creating works that meet the minimum standard of the requirement for getting it registered. 
How would it work? Because even if it were allowed that IPRs may be registered under an AI or that the AI will have the first claim to the creation that would be a dead end. Even if AI is allowed to hold rights, it is still too far in the future to consider AI regulating its creations into the commercial world. Hence, AI systems meeting a certain level of standard requirements should be allowed to get the creation registered under them, just like companies or other organizations. However, the licenses to regulate, duplicate, distribute etc. should lie either with the original creator, or the current owner of the system. Basically, whoever’s made it possible for the AI to create what it did. 
It will definitely require time and detailed research and analysis as a combined effort of technicians and IPR specialists to actually come up with a comprehensive way to effectuate the aforementioned. 

Audi Alteram Partem: Liability
We cannot possibly talk only about the rights to be granted without hearing the other side of it i.e. liabilities. 
AI is programmed with the intention of reducing human intervention as much as possible. This leaves scope for the AI to wander into an unwanted direction. When AI has the ability to create new works it also has the ability to recreate existing works, either intentionally or unintentionally. There is scarce jurisprudence on instances where AI has infringed another individual’s IPR. This brings us to what we had discussed at the beginning regarding AI liability.
The fundamental question here would be the intention of whoever was the most proximate human link to the alleged AI action. Was the person who coded the system meant to code it in a way so as to recreate original works. Or was the person using the AI being aware of its functionality, knowingly fed it data that resulted in the alleged product. 
It would also be worthy to note that what the individual who effectuated the process does with the final product, as it would then show his intentions. If the individual distributes it, sells it, knowingly does so in a dubious way, it would be clear who to pin the liability on. 
If it is not one of the above-mentioned scenarios, it would be better to instruct the user under the guidance of an officially appointed technician to make sure that the AI is modified in a way so as to not recreate the alleged product. 

The ‘Why’ Factor
There have been petitions filed for granting AIs the property rights, discussions, and debates and research on this very topic. But why? 
The main aims of IP rights are promoting innovation and security to creators. However, in specific situations, we may have to look past that and think of the possible conflicts once autonomous creations by machines become a mainstream phenomenon. 
One of the possible points of conflict may be when various minimal-to-no human intervention AI systems promising novelty creations are commercially available. Consumers buying such products will expect the least amount of input from their side and a novel creation. If they do get the desired result, then what would stop the creator of the AI claiming a right to that invention because he is the one who created the AI in the first place. Sure, contracts can be made but how efficient would be signing contracts for every system and honestly when was the last time you actually read the terms and conditions for anything. 
There is already a point of conflict regarding such future regulations wherein ‘open’ movements for science and data are pressing on any and all creations made by AI to be freely available for the development of other AI systems and data. [11] If this is made possible then there would again be no incentive to obtain AI systems that create novelty items as the moment it is created, it becomes public. 
Such problems may be evident now or when we are faced with the situation in front of us. The only way to make sure it does not lead to chaos and hasty decisions is to start regulating it right from the beginning and get a head-start. 

Conclusion: Only the Tip of the Iceberg 
Artificial Intelligence and its capabilities are a gift that mankind has given itself. There are two sides to everything, and AI is no exception. The use of AI should be promoted however while keeping a tab on the developments. Nothing in today’s society can be allowed to stay unregulated. Policymakers have to balance the scales between preserving innovation and curbing misuse. 
AI is still in the process of reaching the stage where it will require compulsory IPR regulation. There are projects under development that will determine the future of AI technology, what we know now is still only the tip of the iceberg. However, one may never be too sure about the possibility of sudden advancement. Hence, the policies must remain in the loop with technology. The earlier we start regulating the larger will be the scope for evolving the policies along with the progression of technology. This will also leave minuscule room for situations that we are completely unprepared for. 
Artificial Intelligence is tricky, as it is an inanimate object with the only quality of humans that we cannot trust as an outsider, their thinking.  

References
[1] The Patents Act, 1970, Acts of Parliament, No. 39, 1970 (India).
[2] Aarsha, Whether Company registered under the Companies Act have corporate personality, Legal Service India, Available here
[3] Bacha F. Guzdar v. Commissioner Of Income-Tax, (1954), AIR 740, (India). 
[4] Politico, Europe Divided over Robot Personhood, POLITICO, (June 25,2020, 23:55), Available here
[6] Ahuja AS. The impact of artificial intelligence in medicine on the future role of the physician. PeerJ. 2019;7:e7702. Published 2019 Oct 4. DOI:10.7717/peerj.7702.
[7] Sandra Durcevic, Intelligent Reporting - How Big Data and AI Are Changing the Way We Create and Analyze Reports, DATAFLOQ.
[8] Bleistein v. Donaldson Lithographing Co. 188 U.S. 239, 23 S. Ct. 298 (1903). 
[9] Eastern Book Company & Ors vs D.B. Modak & Anr, (2007), Appeal (civil)  6472 of 2004, (India). 
[10] Enrique Dans, Meet Bertie, Heliograf And Cyborg, The New Journalists On The Block, Forbes, Available here
[11] WIPO, Artificial intelligence and intellectual property: an interview with Francis Gurry, Available here

0 comments:

Post a Comment