This article has been written by  Priya Jha, a student of BA.L.L.B. (4th year), Bharati Vidyapeeth University, New Law College, Pune.

Abstract

Copyright is meant to secure the rights of the creator of the artistic work. Nowadays, artistic works are produced not only by humans but also by Artificial Intelligence. Artificial Intelligence is a computer system, which is capable of making decisions, having visual perceptions and able to recognize faces and voices like humans do. In short, the computer system has been instilled with human intelligence. Thereby, this also enables the computer system to produce artistic work on its own. Now, the question is that who should be entitled to copyright that computer-generated artistic work, the person who programmed that AI or the AI itself. The law regarding copyright is silent on the matter, as it was formulated with the objective of securing the artistic works produced by the humans or one can say that those artistic work which involves human intervention to any extent. But the ongoing technological advancement demands the policymakers to reconsider the traditional laws.

Keywords: Accountability, Artificial Intelligence, Copyright, Programmers, Work made for Hire.

Introduction

The famous Dutch painter Rembrandt Harmen zoonvan Rijnhas reincarnated almost after 400 years of his death[1]. But this time he has come into this world as a robot named e-David. The robot has been built by the University of Konstanz in Germany.[2] The robot can paint, which doesn’t mean copying already existing works, instead, it takes pictures independently by using the inbuilt camera and draws original paintings with the help of these photographs[3]. So, the robot is capable of producing intellectual property works and it is a better and more advanced version of the Artificial Intelligence (hereinafter AI). It is just an example of Artificial Intelligence. The list doesn’t end here, we have AI Doctors[4], AI Therapist[5], AI Lawyers [6]and the independent driverless car[7]. It has become really hard to discern a field where the AI doesn’t have its influence. It is almost everywhere, be it an investment[8], automated weapons [9]and even espionage   AI is rapidly growing and it is expected to reach the $ 70 billion by the year 2020.[10]Although the purpose of AI is to ease human problems, at the same time it has some drawbacks as well. Among them, one is the challenge of accountability, who will own the products created by AI and who will take the responsibility of the negative outcomes? Further, who will be entitled to copyright the product, who will be reaping the benefits of that copyright and who will save it in case of infringement from others[11]? The article will evaluate many possible scenarios and contemplate the possible drawbacks. After discussing the abovementioned scenarios, The Work Made For Hire (WMFH) will be discussed in detail, as it is the most suitable solution until now. The WMFH solution was proposed by a Professor of Yale Law School, Shlomit Yanisky-Ravid, in an article of Michigan State Law Review.

What is artificial intelligence

Defining the term ‘Artificial Intelligence’ is not an easy job. John McCarthy coined the term “Artificial Intelligence”, but he didn’t define the term. Meanwhile, other scholars gave about ten different definitions.[12] Generally, AI is defined based on their features, like a system which is capable of performing tasks that require human intelligence, such as decision making, creation, learning and evolving[13]. Till now, the AI was capable of performing only those functions, that had been programmed into it by the programmer. It was not autonomous, so, can’t be tagged as intelligent.[14] For instance, a chess engine can play to the extent of its database, it can’t include data itself. Such an AI was not creative, as it would produce only those results that had already been placed into it[15].However, such software has some discretion, which is limited to the operation of the programming of the programmer[16].There was no issue of copyright, as it couldn’t function without any human interference. Merely, works as a tool for human creativity. [17]With time, the AI has been through immense technological advancements. The new type of AI evolved, which can bring some serious problems for Artificial Intelligence through the Lens of Copyright law[18]. The current AI is capable of functioning intelligently and learning things on its own. The quasi-AI learns through hit and trial method, it keeps on hitting until getting the correct response[19].Whereas, the current AI system functions just like a human, it could learn how to paint, compose or write a human does. The products of its creativity are not the outcome of any human input. A creator of the AI has no more claim on the work of AI than the artist’s Mother has on the work of her child.

Who is responsible for the work produced by an AI?

Nowadays, advanced technologies, like AIs, has put up many ethical and legal challenges before us as a society. Such as, who will reap the benefits of the work produced by the AIs? It is not only a question AI unit? Will the programmer be held accountable if an AI breaches into someone else’s copyrighted work? The challenges haven’t been deeply discussed by scholars. In order to answer the above questions, we must see this problem of accountability through the lenses of copyright. There are some problems that have been described by Professor Jack Balkin: the problem of distribution of responsibilities and benefits procured by good deeds or wrongdoings of an AI[20], difficulty arising from the unpredictable character of AI[21], and escape constraints put up to make AI work as per human expectation[22]. In the coming future, we will face the risk of losing control over AIs[23] . Probably we lose control over two or more AI working in consensus behind our backs. So, it is necessary to focus on the issue of accountability.

AI as legal entities

The idea of recognizing AIs as a legal entity has recently been backed up by many scholars. It is due to its autonomy, creativity and spontaneous evolution[24]. In other words, AIs would be responsible for their actions and the outcomes of these actions. It could be done in two ways; first, the personhood approach where an AI would be treated as human. Alternatively, the AI can be deemed as a firm, which is a non-human legal entity.

 Personhood approach

Can an AI be treated as human and given legal rights as humans? According to the personhood approach, the AI systems have the potential to attain consciousness, which means enabling cognitive abilities in robots [25]. The European Parliament Committee on Legal Affairs suggested that the autonomous AIs should be declared “electronic persons”. Hence, they would be responsible for their wrongdoings instead of their programmers. The suggestion was strongly condemned by the experts of the robotics, law, and ethics field, saying that the current technology hadn’t advanced enough to be granted such status[26].However, according to a law professor at Harvard, AI has not only matched the human intelligence but also has surpassed it.[27]

Corporate approach

The corporate approach sees an AI as a legal entity as a corporation, which is subject to labour, corporate and even criminal laws[28].The approach has some consequences, for instance, in Europe, it was argued that robots should pay taxes [29]. It has also been said that they should be punished for doing criminal offences. In this approach the AI systems would be the protagonist; if they are autonomous, then they would be the true owners. Because as a corporation they would have the ownership rights as per the law.[30]

Humans behind robot

Many traditional laws hold this default view of considering the programmers of AI as its owner and thus accountable for it[31]. In the default view, ownership and accountability are given to the programmer of the AI. Sections 14 and 17 of the Copyright Act, 1957 deals with the Meaning of Copyright and First Owner of copyright respectively, which implies that the programmer behind that robot will hold the ownership rights[32].As, the instruments like camera, computer or typewriter are mere tools to produce the resultant work, that too can’t be possible until it is human behind them[33], similarly, the robots can’t function without human intervention. Thus, entities like employers and firms have a claim over the ownership of the copyright as transferees of the programmers [34]. However, AI’s productivity and creativity have gone beyond the reach of traditional laws. They are not able to cope with it anymore [35]. The lawmakers need to revisit the copyright laws as regards the latest developments in AI technology[36].[37]

Distinguish between the rights of AI and the rights of the programmer.

The creators of the traditional artworks were entitled to the copyright of their works. As aforementioned, nowadays the process of creating AI systems involves too many parties. Sometimes, an AI system embedded in the computer generates work which is unknown to the programmer. In that case, the programmer will neither be entitled to the copyright of that work nor of any future works produced by the system. Indeed, the programmer can claim the copyright over the computer itself. AI systems have similar traits of intelligence to humans. Let’s have a look at these traits. First, the programmer of an AI may be unknown to the similarities and interconnections it finds out between things. Second, an AI would grow its own intelligence by using the inputs and its outcomes which is contrary to those fixed and framed computer programs. Third, their trait of self-learning through their experiences makes them unpredictable. So, the traits enable it to produce original and creative work like humans. In short, the claim over the computer doesn’t mean that the programmer would also have a claim over all the products produced by it. Let’s try to understand by taking an example, a person produced a melody over a piano which is created by another person, say

Being the creator of that piano, does X has a right over every melody produced using it? Definitely, X is not entitled to the copyright of the melody, but it is doubtless that he is entitled to the piano itself. Another instance is the well-known selfie taken by a monkey with the camera of a nature photographer. David Slater, a nature photographer, was on the Indonesian island of Sulawesi, there a monkey snatched his camera and took off a selfie. The shutter of the camera was closed by the monkey rather than by David Slater. Alter, the People of Ethical Treatment of Animals (PETA) approached the court on behalf of the monkey. They argue that the monkey holds the copyright of the photo and offered to administer it for the monkey. In the meantime, the David Slater was having the British copyright, which he argued to be recognized the world over. He further pleaded the court to dismiss the PETA’s claim. Hence, the derivative work from the copyrighted program doesn’t let one make claim over the products just because he holds the copyright of the generating copyrighted program. The software programmers and companies in charge merely provide a platform to generate work, it doesn’t make them suitable to be a creator. The producer of the work is entitled to the copyright.

Work suggested

The Work Made for Hire in cases of AI systems

Professor Shlomit Yanisky-Ravid proposed that AI should be treated as a worker to the person who commissioned the work, hence that person would be entitled to the copyright of that product. In short, the professor has proposed to implement the doctrine of Work Made For Hire (WMFH) in the cases of AI systems. Here, the proposal has been reiterated in the essay. The section 17(c) of the Indian Copyright Act, 1957 says that the employer would be the first owner of the work produced by an employee under the employment until an agreement contrary to that has been signed. According to the WMFH doctrine, the employer would have copyright over the works generated by the employee and the independent contractor. The employer can be an individual, a firm or an organization. To ascertain whether a work is made for hire or not, one has to find out whether the work has been prepared by an employee or an independent contractor. If the work is created by an employee, it’s generally considered as a work made for hire. The doctrine of WMFH is an exception to the general rule that the creator of the work would hold the copyright law.

Conclusion

Technology has been growing immensely. Now, we have robots which possess human-like features. These robots have been embedded with programming which allows them to grow on their own. Such features aren’t found in the old model computers which were fixed and framed. The extent of their operation is limited to the programming embedded in them. Whereas, these new model computers are embedded with artificial intelligence. The most astonishing thing about artificial intelligence is that they have human-like features i.e. autonomous, creativity and self-growth. They can create artistic work on their own. The artificial intelligence robots like e-David and Ava, in the movies Ex-Machina are shattering down our preconceived notion that only humans can be creative. However, the artificial intelligence is definitely a boon but has caused a problem in the copyright regime. Who will be accountable for the works produced by these AI? Will it be copyrightable? We have discussed these questions in detail already. The solution proposed is that the AI should be treated as an employee or independent contractor to the user. It is justified by the incentive theory of economics. The user is being incentivized for taking the financial risk of buying in hope of using it for producing. The model has been named as “Work Made For Hire”. The implementation of the mode demands amendments in the traditional copyright laws, owing to its limited scope. Considering the advent of AI in the copyright regime, the policymakers should get alerted and work out the problem soon. As the future lies in the AI, ignoring `it would do no good.

Endnotes

[1] Steve Schlackman, The Next Rembrandt: Who Holds the Copyright

in Computer Generated Art, ART L.J. (June 5, 2019), http://artlaw

journal.com/thenext-rembrandt-who-holds-the-copyright-in-

[2] Jason Falconer, e-David the Robot Painter Excels in Numerous

Styles, NEW ATLAS (June 7, 2019), http://newatlas.com/edavid-

robot-artist-painter/28310/.

[3] Supra note 1

[4] Jolene Creighton, AI Saves Woman’s Life by Identifying Her

Disease When Other Methods (Human) Failed, FUTURISM (June

1, 2019), http://futurism.com/ai-saves-womans-life-by-identifying- her-disease-when-other-methods-humans-failed.

[5]  Jonathan Amos, Love Lab Predicts Marital Outcome, BBC NEWS

(May 28, 2019, 9:20 AM), http://news.bbc.co.uk/2/hi/science

/nature/3484981.stm.

[6]Jessica Chasmar, Stanford Student’s Robot Lawyer has Beaten

160,000 Parking Tickets, WASH. TIMES (June 29, 2016),

http://www.washingtontimes.com/news/2016/jun/29/joshua-

[7] David Szondy, University of Oxford Develops Low-Cost Self

Driving Car System, NEW ATLAS (Feb. 18, 2013), http://

newatlas.com/oxford-robotcar/26282.

[8] Nizan Geslevich Packin &Yafit Lexv-Aretz, Big Data and Social

Netbanks: Are You Ready To Replace Your Bank?, 53 HOUS. L.

REV. 1211, 1222 (2016).

[9] Rebecca Crootof, The Killer Robots are Here: Legal and Policy

Implications, 36 CARDOZO L. REV. 1837, 1840-43, 1863-71,

1894-1901 (2015).

[10] Tech CEOs Declare This the Era of Artificial Intelligence, FORTUNE (June 3, 2016), http://fortune.com/2016/06/03/tech-

[11] Supra note 1.

[12] Stuart J. Russell & Peter Norvig, Aritificial Intelligence: A Modern

Approach, 2-14, 1034 (3d ed. 2010).

[13]Artificial Intelligence, OXFORD DICTIONARY, https://en.

oxforddictionaries.com/definition/artificial_intelligence (last

visited Jan15, 2018).

[14] Arthur R. Miller, Copyright Protection for Computer Programs,

Databases, and Computer-Generated Works: Is Anything New

Since CONTU?, 106 HARV. L. REV. 977, 1038-39 (1993).

[15] Id.

[16]. e-DavId. A Painting Process, UNIVERSITÄT KONSTANZ (Apr.

24, 2017) https://cms.uni-konstanz.de/informatik/edavid/news.

[17] Supra note 15.

[18] Dana S. Rao, Neural Networks: Here, There, and Everywhere—An

Examination of Available Intellectual Property Protection for

Neural Networks in Europe and the United States, 30 GEO.

WASH. J. INT’L L. & ECON. 509, 509 (1997).

[19] Id.

[20] Id.at 46, 48-49.

[21] Id. 45-46.

[22] Id. at 46.

[23] Research Priorities for Robust and Beneficial Artificial

Intelligence, FUTURE OF LIFE INST., https://futureoflife.org/ai-openletter

[24] Samir Chopra & Laurence F. White, A Legal Theory for Autonomous Artificial Agents 1-3(2011)

[25] James A. Reggia, The Rise of Machine Consciousness: Studying Consciousness with Computational Models, 44 NEURAL NETWORKS 112, 112-31 (2013).

[26] Kristin Mangenello, Defining Personhood in the Age of AI, THOMAS ( visited 19th June, 2019). https://www.thomasnet. com/insights/defining-personhood-in-the–age-of-ai/ age-of-ai/.

[27] Glenn Cohen, Should We Grant AI Moral and Legal Personhood?, ARTIFICIAL BRAIN (Sept. 24, 2016), http://artificialbrain.xyz/should-we-grant-aimoral-and-legal-perpersonhoodpersonhood

personhoodhttp://artificialbrain.xyz/should-we-grant-aimoral-and-

legal-personhood.

[28] Brent Fisse & John Braithwaite, The Allocation of Responsibility

for Corporate Crime: Individualism, Collectivism and

Accountability, 11 SYDNEY L. REV. 468, 469 (1988).

[29] Chris Weller, Bill Gates Says Robots That Take Your Job Should

Pay Taxes, BUS. INSIDER (Feb. 17, 2017, 9:57 AM), http://www.

businessinsider.com/bill-gates-robots-pay-taxes-2017-2.

[30] Mark Fischer, Are Copyrighted Works Only by and for Humans?

The Copyright Planet of the Apes and Robots, DUANE MORRIS

BLOG, https://blogs.duanemorris.com/newmedialaw/2014/08/18/

Are copyrighted-works-only-by-and-for-humans-the-copyright-

[31] Robert C. Denicola, Ex Machina: Copyright Protection for

Computer-Generated Works, 69 RUTGERS U. L. REV. 251, 265,

271, 275 (2016).

[32] The Copyright Act, 1957, No.14, Acts of Parliament, 1957(India).

[33] The Copyright Act, 1957, No.14, Acts of Parliament, 1957(India). 34. U.S., Final Report of The National Commission on New Technological Uses of Copyrighted Works 3, 4 (1978),

http://www.digital-lawonline.info/CONTU/PDF /index.html.

[34] Copyright Ownership: Who Wants What?, STAN.U.LIBR,

https://fairuse.stanford.edu/overview/faqs/copyright-ownership/.

[35] Supra note 32.

[36] Id.

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.