The use of artificial intelligence (AI) is making great strides. From assisting in basic household chores to making weighty decisions at the governmental level, AI is being utilized increasingly in many aspects of human life.
With the development of AI, come questions regarding application of the law to AI, such as Intellectual Property (IP) issues related to works created using AI.
At a very basic level, the query is: what is AI? According to the UNESCO, AI is an information processing technology that contains models and algorithms that have the capacity to learn and perform cognitive tasks that lead to predictions and decision making. It follows several methods such as machine learning, machine reasoning and cyber-physical systems.
A basic IP issue arises when AI is used to create a non-tangible product or work that is protected by IP, without any, or minimal human intervention. This scenario gives rise to questions such as whether such work is entitled to any protection, whether AI can be treated at par with a human being, how must the utilization of such work be done, sharing of benefits etc.
A threshold question is whether AI can even be considered as a creator/author of IP. It is settled law that IP is a creation of intellect or of the mind. However, considering the advancement of technology, it is doubtful whether restricting such intellect to humans is correct. It is true that every kind of IP has an end goal of holistic human development. Therefore, providing authorship identity to humans for a work created by AI may lead to the lack of motivation for the human counterpart.
Hence, creating a legal sphere in which AI may also hold some of the intellectual property rights may help us to understand solutions to this question. Even though certain statutory mandates [For eg. PRACTICE AND PROCEDURE MANUAL 2018, Form –XIV: SOP Col 7, which states that the author of a work for which the copyright registration is being sought has to be a human; see also Compendium of US Copyright Office Practices, section 313.2] and judicial observations (Rupendra Kashyap v. Jiwan Publishing House Pvt Ltd, Tech Plus Media Private Ltd. v. Jyoti Janda where courts have observed that there is necessity of human element to be present to obtain copyright; Asia Pacific Publishing Pte Ltd v. Pioneers & Leaders [Publishers]) to an extent, create hindrances in this aspect, the idea of extending personality concepts, as mentioned in relevant statutes, is not theoretically new. Courts have observed that if a conclusion can be drawn, on the basis of cost-benefit analysis, that the larger public is gaining from providing personality and rights associated to an inanimate object, it may be done as such. Further, various authorities have also opined that in the event ‘legislative animation’ is being created, even AI can be deemed to be the creator of IP. Therefore, extending the concept of personality to AI through legal clarification will be the first step towards answering the question of authorship/inventorship of IP.
Jurisdictions around the world are showing a positive inclination towards the acceptance of IP created by AI in equivalence to that created by humans. A recent step taken by South Africa by granting a patent to an invention by DABUS and the Copyright Office of India considering an AI as a co-author of a painting are good beginnings to the trend of allowing rights to subsist in creations of AI. Further, a Federal Court in Australia has also observed that the law in Australia permits AI to be an inventor, but not an applicant or guarantor. China has, both through legislative steps and judicial observations, shown its willingness to accept creations of AI as being capable of protection under IP laws. Here, economic interests of people involved were given substantial weightage in taking such initiatives.
The European Union Parliament had initially adopted a resolution in which it observed that creations of AI should be protected so that interpreters of such work can benefit and obtain rights from it. However, the statutory proposition based on the resolution failed to look into the question of intellectual property rights in creations of AI. Recently, India has also shown a positive inclination towards clarifying the stance on creations of AI. The Parliamentary Standing Committee on Commerce’s report submitted to the Rajya Sabha stated that providing IP rights to works created by AI may benefit and act as an incentive to research and development in the country.
While it is progressive to think of providing protection to creations of AI under IP laws, there are multiple issues that may be a block to the effective usage of such rights. Let us assume that, on the basis of cost-benefit analysis, IP rights are granted to AI, then arises the question as to how these rights can be effectively utilized. How can such IP rights be transferred? What is the role of humans in such transactions? How would the remuneration obtained through use of the IP be distributed? How would infringement of third-party rights be handled? Can AI be made liable for infringement? Answers to these threshold questions are necessary before doing anything else.
In our opinion, one of the ways to get over the initial hiccups would be to consider a ‘guardian’, who may act on behalf of the AI, in order to utilize the IP. A system of automatic assignment can be considered whereby the guardian takes an assignment of the IP created by AI. Once such assignment is done, the attached liability should be also transferred in order to handle instances of infringement, passing off etc. For court proceedings, perhaps, it is possible to apply the principles applied to suits involving inanimate objects like ships, companies etc.
Further, in order to maintain a balance in the situation, law makers also can consider different time periods of protection for the works created by human beings and AI. Lastly, special rights such as moral rights have certain intentions that are exclusively associated with development of human life and hence should not be considered with work created by AI. The rights should be limited to a certain time and no separate rights should be considered than from what is transferred from an AI system.
In conclusion, recognition of IP in works created by AI is a positive approach. It will be an incentive for innovators, but, at the same time, it is important to keep the motivation alive amongst human creators. Determination of such rights should be only viewed from an economic point and should be utilized in a manner so that the person who invest the highest stake will get the highest benefit.
Sujata Chaudhri is the Founder and Alvin Antony is an Associate at Sujata Chaudhri IP Attorneys.