AI

This article is written by Swayamsiddha Das, pursuing a Diploma in Intellectual Property, Media & Entertainment Laws from LawSikho.

Introduction

IP rights are considered to be very different from other laws as it protects an individual’s intellect in terms of creativity, innovative ideas expressed and originality. There are several types of IP rights like copyright, patent, etc. that provide an individual with exclusive rights on the works which have the above-mentioned elements.

In the wake of globalization, every individual wants to be one step ahead of everyone in terms of originality, creativity, and innovation, and therefore the use of technology and software comes into play as people have started using it to achieve their desired goals with a flick of the finger.

Download Now

Talking about technology in the 21st century AI is a term that is heard everywhere. If you are someone who enjoys Wall-E, Terminator, and Interstellar, and you must have been intrigued with how robots were able to act, make choices, build objects, and think like humans?

This very question has now turned into a serious question in the field of IP. As AI has evolved itself, it has started to play a very important role in the generation of new works in the various literary, musical, artistic fields as well as started to innovate new inventions without the interference of humans.

There has been a lot of technical progress in the field of AI over the years, which now helps the computer to learn and execute different tasks on its own without human interference. Having said that, the next question is, while AI has evolved rapidly and is now thought to be capable of producing new works without human interference, are these works open to intellectual property protection? If an AI today makes an original work/innovates a product, can these works be protected by intellectual property laws such as copyright and patent? If this is the case, are the existing intellectual property laws sufficient to safeguard the works/innovations produced by AI? Will the machine be called the author/owner/inventor of these works?

These are the complex questions one could take into account while working with work generated by AI and IP laws. Currently, IP laws all over the world are solely committed to delivering protection to actual humans; however, as AI has attempted to spread its tentacles over the IP sector, it needs to be seen whether there is a scope of consideration or not. Being that, we are seeing how important a role IP can play in the field of AI.

Copyright and AI generated works

Copyright protects works that are original and has an element of the human mind and intellect. Copyright generally protects the expression of the creative and original work generated by the author and not mere ideas.

According to Section 2(d)(vi) of the Copyright Act, 1957 works created by computer and further provides that the person who is responsible to cause the work is the “author”. This clause deals with the concept of AI-produced work with human interference, and it is self-evident that if there is human intervention in the work generated, the programmer retains ownership. Ex: Grammarly is an AI-driven writing assistant tool, the suggestions, changes it prescribes the user to make to improve their write-ups through machine learning are nothing but a case of AI being used to enhance the work while the result is completely based on human effort and intervention. So in this case the author of the literary work is going to be the writer and not Grammarly.

But, what will happen in a case where AI independently creates work on its own, like any other human/individual free from human involvement during the creative process? Will that work be also protected under the Copyright Statute? There are a lot of issues that are needed to be addressed before answering this. The issues are:

  • Whether AI is capable of being an author? 

Authorship is considered to be the most important term in the field of copyright. As mentioned above in the case of AI-assisted work the programmer is given the exclusive right of copyright to do/or authorize a person (to make an adaptation, translation, reproduce, etc.) in respect of the work created under Section 14.

The laws are completely silent as to who is the author of the works generated by AI without human involvement. But if we are considering this then the following can be the approach under the current regime i.e.:

  1. The author (natural person) of the said work may be given the authorship, as AI is dependent on humans for instructions/algorithms. 
  2. It can be assumed that AI was created/programmed in such a manner that AI can identify, encrypt and decrypt inputs on its own if such is the case then the authorship can be given to the developer of such a program as the AI was an intellect of his mind.
  3. The last thing that can happen is either the concept of AI-generated work getting authorship is completely rejected by the state.

Hence the answer to this is NO, AI is not capable of becoming the author of the work as the law currently only recognizes a natural person to be considered as the author of the copyrighted work. 

  • Will the works created by AI be considered to be original?

The most important yardstick for a work to qualify for copyright is “Originality”. Even if the copyright laws have not defined these terms specifically but under Section-13 it is very relevant that for literary, dramatic, and artistic work to qualify for originality the said work must be original. This has also been interpreted by courts in various judgments and precedents.

To check originality there are various doctrines that the court looks into:

  1. The sweat of the Brow Doctrine
  2. Modicum of Creativity
  3. Skill and Judgement Test

For AI to claim for authorship it is required to pass these test of originality, but then however if looked closely AI even if creates work without human involvement would still be at a losing end as it compiles, modifies the works that are out there in the public domain, plus in terms of skill and judgment it lacks skills. Therefore, AI-generated works cannot be considered original.

In the case of Bleistein v. Donaldson Lithographing Co, the court differentiated between the works of a human from that of something artificial. Justice Holmes stressed human nature as an important aspect to create a copyrightable work.  The Court emphasized on the matter that there was no scope for anything that was not a product of man’s creativity to be copyrighted. 

Still, there is a ray of hope for the works generated by AI to be considered as original as discussed in the case of Alfred Bell & Co. v. Catalda Fine Arts, The District Court of the U.S, lowered the requirement for originality, ruling that it could not be replicated from any other artistic work of a similar character to be considered original. The judgment was favorable for claiming copyrights of works created by AI because such works were not copied but extracted by programming and algorithms. 

Taking into account both traditional as well as the modern/practical views we can form an opinion saying that the laws are somewhere trying to match the pace with the current requirements and is trying to deal with these technical aspects revolving around AI and IP protection.

  • Can AI be considered as a Legal Entity?

Addressing the issue of AI is considered a legal entity to answer this. We need to understand the fact that AI is always seen as a tool that assists humans in getting their job done. The laws do consider the artificial person like a partnership, companies, etc. as a legal entity and grant copyright to them in works generated by them, provided, only if it proved that they work and function with the help of human involvement. For this very reason AI even being considered as an artificial person is still considered as a legal entity. Having said that it also makes answers the issue of who is liable in case of infringement caused by AI? having said that AI is not a legal entity it becomes very difficult to place the liability of infringement as AI lacks legal status, hence any infringement caused by it would become a serious issue.

Therefore, to maintain the balance between copyright laws and AI it is very crucial to address the loopholes surrounding it and amending the laws accordingly. And the concept of authorship and a legal person should be changed as per the need of time.

Patent and AI generated works

In India, patents can only be obtained if your invention is unique and new. Over the past years, there have been various cases where inventors have filed a patent application for AI-related inventions, India has been no exception to these kinds of applications. Normally the patent law grants exclusive rights to the inventor which excludes others to use, make and sell a patented invention for a limited period. To make your invention patentable, your invention should have 3 things:

  1. The invention should be novel.
  2. The invention must be non-obvious i.e. there must be a significant change in the invention as compared to the previous one.
  3. The invention should be useful for mankind.

Coming to inventions invented by weak AI (like Alexa and Siri who can understand your commands like set an alarm, play songs, turn on the lights, etc. but it doesn’t comprehend any meaning of what you said) it is not ambiguous as in who will be considered as an inventor, but problems and questions arise when inventions invented by Strong AI (AI programs that can surpass human intelligence i.e. superintelligence) in these kinds of situation ambiguity arises whether inventions invented by the Strong AI technology should be patented or not? If AI can be considered to be an Inventor or not? 

At the present, the answer to these kinds of situations is NO as, under patent law only a ‘natural person’ can be identified as the inventor and only a ‘natural person’ can be granted a patent. If the inventor is a Natural person who serves the purpose of getting a patent in any situation, then Sophia the AI Robot who has been granted citizenship in Saudi Arabia would be considered as a Natural Person?

As discussed above an invention can be patented on the ground of non-obviousness does this mean that Stephen Thaler’s AI machine DABUS (sole inventor of 2 discoveries in the field of medicine and drug) would not be considered as the inventor merely because it isn’t a natural person? but then if looked into the provision of the Act, Section 2(1)(p) defines patentee as the person who has entered the patent office as the grantee or the owner of the patent, from this section it is safe to draw that the inventor is ought to be a natural person, but then according to Section 2(1)(s) the definition of person includes both government and a non-natural entity, but According to Section-6, the true and first inventor has an exclusionary definition and Section 2(1)(y) doesn’t mention the need of natural person regarding an invention invented in India. Although these provisions do not explicitly state that an inventor must be a natural individual, it suggests that human interference is required for an invention to be deemed patentable. Hence it can be seen that even if it is established that inventions made by Strong AI are non-obvious and unique it is still unable to uphold its rights as the laws are not yet ready for these drastic changes. 

Apart from this, the major issue that needs to be addressed is in terms of liability and ownership, as into who will be held liable in case any sought of illegal action is brought against the action of AI, is it the software that is to be held liable or it’s the owner of the software? And does the AI have the power to assign or transfer ownership to the 3rd parties?

To deal with these kinds of the situation at the present we consider the 3 principles laid down by legal scholar Gabriel Hallevy:

  1. “Perpetration-via-another liability model”– this gives the situation where mens-rea (the knowledge of the wrongdoing) would not be applicable on the AI concerning the illegal activities rather the perpetrator would be either the owner of the software or the user of the AI.
  2. “Natural-Probable-Consequence liability model”- this normally holds the programmer/ users responsible for the illegal acts committed, as it believes that the programmers or users are deeply involved in the everyday operations of the AI entity, but without intent to commit an offense. But then as it is known that ignorance of the law cannot be used as a defense, it assumes that the programmers or users of an AI should have known about the likelihood of committing the relevant offense, and therefore holds them responsible.
  3. “Direct Liability model”- this suggests that AI should be entirely held responsible for the illegal acts committed and not the programmer/user.

There has been a significant change that is waiting ahead of us, currently, the laws are not clear and are ambiguous on certain terms and provisions but because the future lies in AI the laws need to adapt to the new AI dominated world and find answers to the issues at the earliest.

Designs and AI generated work

The main purpose of Design rights is to protect the aesthetic element of a product, these aesthetic elements might include shape, color, pattern, etc. Industrial Design may be either 3D or 2D in nature. As we know that AI technology is capable of generating Artistic work they are also capable of creating 2D and 3D features of the concerned product. 

Design created by AI technology, whether with human intervention/without human intervention are termed as “AI-generated designs” but then again the similar issues regarding Whether AI-generated design can be granted design protection? Whether there should be any kind of special legal framework that deals with the authorship and ownership aspect in the case of AI-generated work?

Currently, the laws are silent about these issues and there is no concrete answer to this question. Normally there are 2 types of protection under the design rights i.e. registered and unregistered designs. In cases where AI-generated wants seek protection under registered design, the chances of getting it might narrow down as issues like identifying the author and owner are still ambiguous as it is assessed individually solely based on the merits of the case. 

At present, the law has not addressed the elephant in the room and is therefore open for various interpretations. The stringent laws and provisions need to change and should be at par in the tech world.

Conclusion

The present situation surrounding AI and IPR is very challenging as there are a lot of identified grey areas surrounding this very issue of AI-generated works. Currently, the IP regime is not ready to address this.

At present we are only relying upon precedents and interpretation of the courts, but in the long run, this might not work. There is a dire need to create separate sections and provisions that can deal with AI-generated work as well as introduce special laws that deal with the liabilities and remedies in case of infringement caused. Various international conventions like WIPO have already considered the above matter and have discussed the pros and cons of the same but still, there is a requirement for a uniform structured policy at the international level.

But then, because AI is a relatively new topic so there is no explicit provision as of now but it is taking the words by storm so the existing provisions must be interpreted in such a way that we can reap the benefits of technology without jeopardizing our rights.

References


Students of Lawsikho courses regularly produce writing assignments and work on practical exercises as a part of their coursework and develop themselves in real-life practical skill.

LawSikho has created a telegram group for exchanging legal knowledge, referrals and various opportunities. You can click on this link and join:

Follow us on Instagram and subscribe to our YouTube channel for more amazing legal content.

LEAVE A REPLY

Please enter your comment!
Please enter your name here