University

School of Law

The Copyright Protection of AI-Created works by the European Union Copyright Legislation

Date of Submission
Table of Contents
Abstract 4
INTRODUCTION 5
Background Information on Artificial Intelligence and its Relation To Copyright Legislation 7
Statement of the Problem 10
Research Questions 11
Significance of the Research 12
The Structure of the Research 12
LITERATURE REVIEW 13
Research Question 1: 13
Originality in the Acquis Communautaire 15
Research Question 2: 19
Research Question 3: 20
RESEARCH METHODOLOGY 25
RESEARCH DISCUSSIONS 26
Research Question 1: 26
Research Question 2: 30
Research Question 3: 35
CONCLUSION 38
Bibliography 39

Abstract
Copyright Law in the European Union has increasingly focussed on increasing the scope of works that have a right to intellectual property protection. Currently, the law covers various work categories such as literary works, music, film, sound recordings among others. However, the European Copyright Legislation framework does not consider non-human innovations yet artificial intelligence advancements continue to lead to the development of machine-generated creative works. Currently, there are developments that allow autonomous programs to create works attached with high value which range from software to literary works, images and music. The creative process to these works will have limited or a non-existent human contribution. Therefore, creating a need for improvements in the Copyright Regulations so as to adequately canvases all the AI-created works.
To this effect, this research paper undertook a critical analysis of the present EU Copyright legislation to understand its position when it comes to copyright protection for the AI-generated works. Additionally, the research answers why the AI-generated works and to be protected and which are the legal tools that could be improvised to grant copyright protection. The famous legal dogmatic approach was incorporated into the research where various scholarly material from peer-reviewed sources are utilized in gaining insight on the research subject. In conclusion, the researcher does acknowledge the lack of clarity within EU Copyright’s Legislation in relation to AI-created works. Additionally, the research indicates that protection of these works should ensue considering it is an incentive to the developers and also a guarantee of technological development to the entire society.
INTRODUCTION
Copyright Law in the European Union has increasingly focussed on increasing the scope of works that have a right to intellectual property protection. Currently, the law covers various work categories such as literary works, music, film, sound recordings among others. Computer programming works such as Artificial Intelligence (AI) are considered to fall under the category of literary works. According to Section 11(1) of the Copyright Designs and Patents Act (CDPA), the author of the literary works is the initial owner of the copyright title and the rights affiliated with the title. The author should have created the work and the after need to be original” exhibiting a particular degree of labor, skill or judgment. Currently, most litigation cases need the input of the human authors to determine the originality of the works.
Notably, the intellectual property (IP) rights granted to AI have continually generated various issues whether legal, economic or moral. This is because the AI technologies have been developed to act autonomously without the input of persons. Conversely, Section 1 (a) of the CDPA has indicated that any work that has no human author is not satisfactory enough to be a subject under copyright law. The UK Copyright law has tried to cover up this print through Section 178 which acknowledges “computer-generated “ works specifically those generated in circumstances where there is no involvement of a human author. Section 9(3) of the CDPA has indicated that the author of these computer-generated works will be the individual who made the necessary arrangements in creating the works. There is a new legal framework that has provided copyright protection to the artworks created from AI.
Furthermore, the European Union (EU) is also focussed on being the most advanced region when it comes to AI. The European Commission made an announcement on 25 April 2018 in relation to various measures in putting AI to utilization within the region and boosting its competitiveness. The Commission indicated that it does acknowledge the challenges faced in trying to make AI succeed and work for each individual. Nonetheless, it is ready to face them which began with an investment of at least 20 billion euros by the end of 2020 for further research to develop AI technologies, applications and have companies embrace them. This development objective by the EU also requires legislative clarification since the legislative environment has been noted to have a negative effect on investments and the ultimate societal development.
The European framework of copyright law does not consider non-human innovations yet artificial intelligence advancements continue to lead to the development of machine-generated creative works. Currently, there are developments that allow autonomous programs to create works attached with high value which range from software to literary works, images and music. The creative process to these works will have limited or a non-existent human contribution. It is prudent to state that the creation involving artificial intelligence has a fundamental difference to the traditionally creative processes. Therefore, applying the existing copyright legislation to protecting the AI-generating works is already a novel challenge. Copyright protection of AI generated works is a challenge that the EU faces. Various institutions in the region have explicitly addressed the need to kandle the legal implications affiliated with the developments of Artificial Intelligence. The lack of clarity to copyright legislation of AI generated works grants its developers no incentive to engage in further development of valuable AI systems nor are the companies willing to invest in the technologies. The EU will only maintain competitiveness in the global market by clarifying on the copyright legislation for AI-generated works.
The central aim objective of this thesis is undertaking an analysis on whether the AI-created works have been adequately canvassed by the European Union Copyright legislation and also evaluate if they need further protection through an apportionment of rights. Ultimately, the thesis will also determine the proper legal tool to provide protection to the AI-generated works. The research will incorporate the famous legal dogmatic approach of systematizing and interpreting the existing respective law content. The method also seeks to provide justifications ypo interpreting the existing law in relation to the research’s subject of finding whether AI-created works are fit for protection under the current EU copyright legislation. The target is to find the answer through an interpretation of existing legislation and case law. Furthermore, the legal histories, copyright directives, legal provisions, academic articles and international treaties will be considered to gain more insight on the originality and human author requirements stipulated in the EU.
Background Information on Artificial Intelligence and its Relation To Copyright Legislation
AI is a field of science and set of computational technologies which are influenced by the manner in which individuals utilize their nervous systems and entire bodies for sensing, learning, reasoning and taking action. The main objective of AI research is simulating human intelligence to the extent of having programs acting and reasoning accurately and autonomously. An autonomous system will be one with an ability to carry out an assigned activity without guidance from human beings. The AI concept continues to be a discussed topic with respective scholars not reaching a universal agreement. Nonetheless, Nils J. Nisson provides a definition that has been accepted by a huge percentage of the academic community. He stated that AI is the activity that is devoted to making machines intelligent with intelligence being a quality that enables an entity to function properly with foresight of its environment.
The AI field was originally born in the 1950s where researchers initiated investigations into the simulation of machines to have aspects of human intelligence, Alan Turing was acknowledged for having the most influential ideas and also for proposing the formal model of computing. In his essay, Computer Machinery and Intelligence, Turing has imaginations of the probability of having computers with simulated intelligence and also exploring various AI fundamentals. These fundamentals included testing the intelligence,and programming the machines for automatic learning. Over time, there have been heavy investments on the field’s research and development. Despite the challenges faced in the way, fifty years later the researchers state that they have finally achieved the success predicted in the initial years of its development. Currently, there are hardwares that are cheap and more reliable in sensing, actuation robots that can easily build and the internet with large datasets for programs’ training. Furthermore, large computing powers and storage capacities allow statistical techniques for deriving solutions and applications from the gathered information.
In regards to AI being utilized as a tool in creation of works, it is important to note that in reality AI is very far from being really “intelligent” even though it may illustrate equivalent to the human mind. This is because of the distinct definitions to intelligence. If intelligence is seen as the ability to find new ways that are not known then AI could be termed “intelligent” since it has the capacity of detecting new relationships within large data sets which was previously impossible. Additionally, AI could identify previous errors and improve the patterns within the programs. Therefore, some commentators present the argument that the author no longer has substantial control over a particular work that has been created and under distinct conditions. Conversely, the present forms of AI fail to determine the preferences or aims that need to be achieved and the human being implementing it needs to define them. Therefore, AI will improve the ways of achieving an objective but cannot change it. There have been suggestions to consider AI or robots as a new form of legal person (ePerson), its limitations disregards the facts and also the suggestions do not answer critical queries such as the liability of aI and whether it can raise money to deal with claims raised against it. At length, the real question concerning all the legal aspects related to the actions of AI is whether the latter could be attributed to the individual using it.
In copyright legislation, AI raises concerns on whether its created works should be considered as personal intellectual creation which is essential in acknowledging the work’s copyright protection. Since the behavior of AI is more or less uncertain, the traditional and deterministic strategy in dealing with utilizing digital tools is problematic in application. Prior to the development of AI, the utilization of software would easily be attributed to the author since the outcome was principally foreseeable. However, the conditions have changed as the author uses the AI and can only determine the primary preferences and goals to be attained. The situation is almost equivalent to works of art that have been created using software that randomizes the use of distinct colors among others. The results from using AI are unpredictable and some creators have stated that the primary point of their creative process is availed by AI and not the author. Therefore, the author or associated human being is not considered the determinant of the work and the latter cannot be acknowledged as copyrightable work.
Conversely, this perspective does overstate the function of creativity and the range of capabilities of AI. in a legal sense, AI cannot be considered intelligent hence cannot be compared to human will. Since it can also not set its own objectives and preferences, the author that makes AI useful and defines the framework of a particular work remains an important element. Still, there are good arguments that attribute the AI-created works to the person making the technology useful. For instance, if the artist chooses to train the AI specifically on the Rembrandt paintings then AI elucidates a “new Rembrandt”, its extent depends on the degree in which the artist will influence the AI. in case the artist utilized particular paintings and not all then there is a strong influence on creating the ultimate “AI painting”. In such an event, the work should be attributed to the artist. If the artist decided to train the AI depending on all the paintings including other artists’ paintings, then setting the objectives and respective framework will have a considerably lower meaning. If an artist fails to know that AI has been trained, it then becomes difficult to speak on the artist’s creativity.
Another aspect of copyright law is the protection of AI itself. The present legal framework, AI is not being considered as a concept or algorithm that needs protection, it is considered a code as or the EU Software Directive. Furthermore, the data utilized in training the AI has also not been protected. For instance, an AI that depends on a database, the the database’s structure has been protected by the EU Database Directive, the data that has been produced by the AI has however not been protected by copyright law, the new data that has been generated by the AI system could be considered a trade secret as per the new EU Directives dealing with protecting the undisclosed know-how and business information. However the protection is still a weaker front compared to protection under copyright law. Trade secrets are not real property rights.
Statement of the Problem
AI-generated works have been developed constantly to reach a point where their creative processes have minimal or no human input. This creates a challenge for copyright law which has been protecting creations done by human beings. Presently, the EU Copyright Law has not handled the protection of AI-created works especially where the creation process cannot establish a human author. Therefore, when the AI systems have independently created a work of art, there is a question of whether protection should be granted to the new types of artwork. Copyright protection is an IP right which grants a work’s creator exclusive rights to determine whether another party could use their creation and the conditions for the respective usage. The creator is permitted to utilize their works commercially specifically selling, distributing or granting license rights to others. AI-generated works need to have established copyright protection within legislative frameworks to prevent any illegal forms of duplication and replication.
The European Copyright Law is also considered to be ambiguous when it comes to AI-generated works which makes it difficult for the creators to have proper assessments of the value of their creations within a digital environment. This induces scepticism even in engaging and developing new AI technologies. For instance, greater levels of digitization through AI will have the latter accentuate on the matter. Many creators have found themselves in a dilemma whether copyright law can protect their works and what are the associated conditions for the protection. AI is not termed as a legal entity under the EU law and thus cannot demand ownership for the work it has created. Professor Ole-Andreas Rognstad stated that such a situation increases the possibility of a “no ownership scenario” for the respective works which will have the latter falling into public domain. This means that the creators lack the incentive For creating the AI-based programs due to the lack of copyright protection. This slows down the technological evolution that can happen and any affiliated areas such as new enterprises, pharmaceutical organizations among others and finally the consumers will all get affected.
Directive 2009/24/EC has established that the computer program to be protected has to be original in terms of being the author’s individual intellectual creation. There are no other criteria applicable to determine whether one one is eligible for protection. Therefore, work created from a program, has originated from the creative input of the creator thus can gain copyright protection. However, there is still a contentious issue in this perspective considering that AI is continually evolving even through a self-learning procedure to produce autonomous works that have no input from individuals. The creator of the initial computer program is at the time not required by the AI for the creative process. Furthermore, the AI will develop itself and create work in a manner in which the work and the original computer program cannot be considered as original due to the extreme inconsequential association with the creator. Subsequently, a challenge related to the originality condition has been created when there is a breed of granting copyright to the AI-generated works.
The General Director of WIPO, Francis Gurry, stated that AI is a new digital frontier that will have a substantial effect on the world. This makes it an issue when looking into the digitization of music creation or the fact that scholars will use AI such as the 3D printer in the creation or recreation of objects and paintings. To this effect, there is an evident need for further analysis of the European Law. there is a question of whether originality is actually the most proper condition for the AI-generated works. With the digital environment expanding and bringing in new challenges, maybe there needs to eb an additional layer in IP toi cover them. The present EU Copyright law has not offered any criteria for protection especially one that is adequate for these kinds of works. It may be a time that other legal solutions are explored.
Research Questions
1. Have AI-generated works been adequately canvassed by the European Union Copyright Legislation?”
2. Should the AI-generated works be provided with adequate copyright protection and be apportioned the respective rights in the EU?
3. Is there a legal tool that needs to be implemented that will ensure granting of copyright protection rights to the AI-generated works within the EU?
Significance of the Research
Copyright protection rights to AI-generated works constitute a current and difficult challenge which needs further deliberation. The enlisted research questions will look into the current position of the EU Copyright law when it comes to works co-created or independently created. This discussion is expected to demonstrate that the current copyright legislation was formulated with Artificial Intelligence not in consideration. Consequently, applying the traditional regulations to these works only leads to lack of clarity and the unintended solution to some scenarios. This challenge does warrant a discussion of the suitability of the present solutions and also determining whether EU’s copyright legislation has to be adjusted for better regulation of AI-created works. The European Parliament and Commission also acknowledge the need for a resolution for the current IP legislation to be altered to consider the AI advancements.
The Structure of the Research
This first chapter was focussed on making the reader familiar with AI-generated works specifically in the field of copyright legislation within the EU. The subsequent and second chapter is the research’s literature review which will look into various scholarly research that have discussed the EU Copyright legislation. The review will also consider the underlying rationale utilized in granting copyright protection rights and the views of scholars on whether the EU has considered the AI-generated works in therationale. Additionally, the review will also look into any proposed legal tools that could be used to apportion the rights yti these respective works. The third chapter will discuss the research methodology utilized in this research. The fourth chapter will aim to answer all the research questions in-depth. The fifth chapter will be the conclusion of the thesis to the research questions and any recommendations that could be incorporated into EU Copyright legislation in respect.

LITERATURE REVIEW
In this literature review, the researcher is to carry out a critical analysis of research arguments of various scholarly material in relation to the subject of research. The scholarly material selected provided in-depth insight on Copyright protection within the EU Copyright law specifically whether the latter has canvassed the issue adequately and any recommendations they gave to handle any gaps in the regulations.
Research Question One:
Have AI-generated works been adequately canvassed by the European Union Copyright Legislation?”
A substantial percentage of copyright legislation found among the EU members states heavily rely on human-centered aspects. These include having a beneficiary to the protection who is typically the author of the works, existing conditions for the protection to happen such as originality and the granting of rights that are normally economic or moral. The human-centered approach that is common in these copyright laws is also evident in the acquis communautaire even if to a lesser extent for lacking the regulations related to moral rights. Additionally, the Software and Database Directive have defined authorship to copyright works in accordance to natural individuals or groups of natural individuals having created the works. Iglesias et al (2009) indicates that the anthropocentric approach has also been incorporated into defining originality. While the European law has not given proper clarity on the definition of originality, many of the directives within the EU law have linked the quality to natural persons or person attributes. The Resale Directive considers them the persons/artists and the Copyright Term Directive considers the human attributes/personality. Furthermore, the Software and Database Directives together with the Term Directive consider the “author’s own intellectual creation” when speaking of photography and staging that it is the only criterion to be followed in the assessment of originality. The Court of Justice of the European Union has tried to harmonize the subjective dimension through a series of landmark decisions in cases of copyright-protected works. In these decisions, the court indicates that the “author’s intellectual creation”,“The free creative choices”, “the creator’s personality” or the “creator’s personal touch” as the fundamental criteria in determining whether the work should be copyright protected.
An analysis of the various copyright legislation in respect of AI-generated works, Iglesias et al (2009) contends with many European scholars who indicate that the present European and national legislation is not eligible to protect the works through copyright. This is because of the fact that the works that have only been created by the machines cannot be termed as original since the copyright laws clearly indicate that they should have human attributes affiliated with them for copyright protection. Therefore, Iglesias et al (2009) indicates that since the AI-generated works fail to fall under the present copyright legislation in the EU, there is a need to have an assessment done to determine the category they should fall under in order to receive protection. This is a process that will require meticulous examination of the protection’s rationale, any market failures and whether incentives need to be created. The research indicates that the answer chosen to this challenge should also encompass the possible impact of the protection on the market in regards to the creative works and innovation, in case there is a conclusion of having the copyright rights created, thena further consideration on defining and implementing the rights should ensue. Nonetheless, as at the time of Iglesias et al (2009), their research indicated that there was no economic research that had been done on the field despite various legal scholars addressing the subject.
AI has received wide reports on media platforms especially due to its utilization in generating news, composing music, creating artworks and producing scripts. The recent developments that have been achieved by the AI techniques have allowed machines to get to autonomous levels thus trivializing human contribution into the creative or inventive processes. According to Iglesias et al. (2019), particular countries such as the UK, South Africa, Hong Kong, New Zealand among others have set up regulations to avail copyright protection to the individual that set up the necessary requirements in creating the work. In respect to the UK, the computer-generated works are considered to be the works that have been generated by the computers in situations where no human author was involved. It is prudent to note that the provisions provided in the UK leave no room to have ownership allocated to the programmer or the user. Iglesia et al continues to state that case-law related to copyright protection for AI-generated works are scarce. The works are able to enjoy a shorter period (50 years) of protection compared to other copyright-protected works that get 70 years. However, for the countries which do not have particular regimes, concerns are raised on whether AI-generated works actually get copyright protection or whether they would be considered original and the allocation of ownership to the rightful owner.
Some legal scholars have raised opposing perspectives such as Dickenson et al. (2017); to Guadamuz (2017); Lambert (2017); Lauber-Rönsberg and Hetmank 2019. According to Guadamuz, (2017), the approach that has been incorporated by the UK would handle the challenge within the European Union and indicated that the refusal of granting copyright protection to the AI-generated works could result in a severe commercial impact especially in the field of databases. The UK system does exhibit evident advantages such having a particular level of certainty in a very uncertain legal area. Guadamuz (2017) indicates that the system has already received global recognition as various countries implement it, the approach is ambiguous enough for deflecting the user/programmer dichotomy Question Assignment and analyzing it on a zaze by case basis. However, the UK system has also received its own critique from its opponents who consider it not to be entirely appropriate in handling the computer-generated works. The first critique is that the UK provision has hugely emphasized on the difference between the Helped and non-Helped creations. The second critique is that the provisions have left a substantial margin of uncertainty on the individual that made the creation’s arrangements specifically for the highly complex systems. The third critique entails the UK provisions not addressing the originality issue. Another critique is that they fail to hand;e the challenge of the jointly generated creation by human beings and the machines. Finally, there is a question of whether law does comply with the EU acquis.
Originality in the Acquis Communautaire
According to Margoni (2018), the originality standard has historically been an issue in domestic and international level such as the Berne Convention that only needs to work to have an intellectual creation. In the EU, while it does have the originality standard, things would start to change in 1991 after implementation of the first copyright Directive – the Computer programs Directive. This directive harmonised the standard to the level of the author’s own intellectual creation. Nonetheless, as a legal instrument and with the implementation of the Term of Protection and the Database Directives, this standard would only have vertical harmonization whereby only particular subject-matter being regulated by acts stipulated in the EU secondary legislation. The sectorial harmonization done to Copyright legislation also has a possibility of being found in a clear and direct power attribution of powers within the EU in regulating copyright. It is until recently that Articles 26 and 114 of the Treaty on the Functioning of the European Union (TFEU) became the main basis for interventions used within the EU in respect to the copyright field. These articles have given the region the competence to incorporate measures respectively with a target of creating and corroborating the internal markets’ functioning and law approximation in the member states. The lack of clarity in the attribution of powers for copyright regulation resulted in the fragmentary and subject-matter approach that has been adopted by the EU Copyright Directives.
The start of the 21st would have things changing due to the landmark decisions made by the CJEU as they tried to clarify and establish that the originality standard was not only applicable vertically but also horizontally. The court indicated that the author’s own intellectual creation” was not only for software, photograph and databases but extended horizontally to all the subject-matter included in the Information Society Directive or in the Berne Convention while excluding the registered designs. Furthermore, the EUCJ also established several interpretive elements to be utilized in determining whether the work is original. Within the EU, this quality is attained when the creators can partake free and creative decisions while also putting a personal stamp on their creations. Conversely, this is not available when the expression is to be determined via technical or functional regulations such as when there is only a single method in expressing an idea or the expression was previously determined by a particular objective or narrow rule have provided constraints due to the lack of space to partake in free and creative choices. However, the works utilized by the Court should not be termed as the requirement of “author’s own intellectual creation” being at a higher level. In actual fact, an analysis of the facts demonstrates a distinct result. The Court identified a new standard that gives more emphasis on the qualitative type of authorial contributions rather than the quantitative contributions. Nonetheless, its exclusion from the single sentences cannot happen and if original then it can be an object of copyright protection. This outcome is something that is not included in the UK regulation prior to the EU harmonization at least in consideration of the titles. Conclusively, Margoni (2018) indicated that there is an assumption that most corpora used for AI especially those having a literary and scientific chartacted have copyright protection. Furthermore, if parts of the literary works which could at times be very short, as long as they are original in their own right, they are to be considered to be protected and any kind of reproduction reserved.
Notably, Lukoševičienė (2017) does explore the originality requirement that is stipulated by the EU Copyright law during granting copyright protection and its significance in the creative sharing community of WIkipedia. His research is also focussed on contributing to the future of the copyright debate. The researcher does indicate that the “author’s personal touch” and the “free creative choices” brought forth by the CJEU may not be appropriate practically and also how the author has been conceptualized is not aligned with the understanding of a Wikipedia community. The researcher acknowledges the fact that copyright law debate is continually made complex due to the probabilities of present technologies which makes it hard to control distribution, the duplication of copyright material and also obscures the conventional boundaries in the dynamic digital atmosphere. Lukoševičienė (2017) indicates that with E copyright law aspiring to meet the needs of the intrinsically motivated and self-organizing mutual communities, the originality standard is a criterion in awarding protection should highly consider the skill, effort and value of the final product. This is because following the path stipulated in its continental copyright and focussing on the author and the creative process, this results only leaves out the online sharing communities such as those in the Wikipedia space. Technology, community and the final product undergo an integration to a form that is almost inseparable as a whole, the technology’s openness and the aspirations of the community walk hand in hand. The researcher indicated while the originality requirement is substantial, there should be more criterions in determining whether particular works are to receive copyright protection due to the increased varieties of digital tools and products.
As it has been noted, while the law is clear in the UK covering computer-generated works, the situation in the rest of Europe is considerably less favourable towards ownership of computer works. There is no equivalent to s9(3) in the major continental copyright jurisdictions, and the subject is not covered by the international treaties and the copyright directives that harmonise the subject. Art 5 of Spanish copyright law specifically states that the author of a work is the natural person who creates it; while Art 7 of German copyright law says that the “author is the creator of the work”, and while it does not specify that this is to be a person. Article 11 declares that copyright “protects the author in his intellectual and personal relationships to the work”, which strongly implies a necessary connection with personhood. The end result is that computer-generated works are not dealt with directly in most European legislation, so when presented with a work that has been created with a computer, one must revert to the basics of awarding copyright protection, namely originality.
For such a vital concept of authorship, originality has proved to be a difficult concept to pin down, while it is well understood that originality is one of the most important elements of authorship, different jurisdictions have developed their own version of originality, and Furthermore, the level of originality may vary in one jurisdiction depending on the nature of the work. Indicative of the lack of harmonisation is the fact that Rosati identifies at least four different originality standards in common use. It is precisely the European standard that could present its own unique challenges to computer generated works.
Research Question 2:
Should the AI-generated works be provided with adequate Copyright Protection and be Apportioned the respective rights in the EU?
There have also been other scholars such as Ramalho (2017) who have indicated that there is a need for incentives for creating or commercializing AI-generated works. While many of proponents of this argument have not fully supported copyright protection for the AI-created works, they indicate that there is a need for an alternative protection to these works. Ramalho (2017) suggested a solution in accordance to the public-domain model in conjunction with a “disseminator Right ” which has been highly influenced by protection given to the publishers of unpublished works falling under the category in the Copyright Term Directive. For instance in the United States Samuelson (2985) indicated that ownership to computer-generated works need to be allocated to the users. Other scholars such as Bridy (2016) and Pearlman (2018) stated that implementing the work-for-hire doctrine of the AI systems. Ginsburg (2018) suggested giving the sui generis or neighbouring right which considers the potential option in assessment. In case of proof of creating a new right, the influence should be taken from the industry property rights which are distinct from copyright and a registration process has to be followed.
On the other hand, some authors have stated that there is no empirical evidence that supports the need for property rights for AI-generated systems. Pierry & Margoni (2018) indicated that the AI-generated works should be excluded from IP protection rights and have them fall under the public domain. Others such as Bently (2018) stated that rather than grant the IP rights, the contracts should be considered the unfair competition law. While there is no empirical evidence on what would be the impact of having the AI-generated works fall under the public domain, one notable impact is that it has a negative effect on investments. In case developers have doubts on the creation developed by AI qualifying copyright protection, then why should there be an incentive to put in investments into the systems.
Notably, in determining whether the AI-generated works need to receive copyright protection, there are evident impacts in choosing to grant or deny the works copyright protection rights. According to Michaux (2018), there are challenges in differentiating the works generated by individuals and the AI-generated systems. Even while trying to deprive the latter of copyright protection. This challenge world still prevails even if the protection was given based on a distinct right. Conversely, Michaux does predict that the protection could cause a considerable increment in the number of protected works and the copyright concentration within a small number of corporations. The protection could cause an instrumentalization of the work notion and increase the cost to accessing various works.
Moreover, Lauber-Ronsberg (2019) does express similar concerns indicating that granting copyright protection to the AI-generated works will extensively derange the principles of copyright law and the tenets upon which this legislation system was built. The researcher acknowledges the fact that creating an ad hoc right may be a better option to dealing with the uncertainties of AI. However, she also has her concerns with the requirements needed to grant the protection and allocating respective rights. Furthermore, another challenge is the mandatory or voluntary closing of the utilization of AI in a particular work or related work which can be very challenging for implementation. According to Perry and Morgan (2010), there is a problem with constantly eroding the public domain while Schonberger (2018) indicates the possibility of incentive destruction for the human authors in case the AI-created works receive similar treatment as the former.
Research Question 3:
Is there a legal tool that needs to be implemented that will ensure granting of copyright protection rights to AI-generated works within the EU?
According to Ciani (2019), even with proper legal standings on copyright protection to the AI-generated works, there are still legal provisions that are governing the field. The lack of legal standings requires the main stakeholders to first determine the authors of the AI-generated works. For the arutoi case, Judge Orrick gave no clarifications on the questions. Nonetheless, some law jurisdictions that could have guided the process including the U.K. Copyright, Designs and Patents Act, 1988 [§ 9(3)], Irish Copyright and Related Rights Act No. 28/2000 [§2(1)] and New Zealand’s Copyright Act of 1994 (§ 2. These regulations have vested copyright in the works that have been generated by a computer ibn circumstances where a human author is not present to the individual that made the necessary arrangements in creating the work. According to Section 12(3) of the UK 1988 ACT, it considered the protection term in the AI generated works to last 70 years from the year in which the creation.
Nonetheless, there are increased doubts where the provisions are considered to be covering situations especially during the end work as an autonomous creation. Furthermore, there is a debate on what are the arrangements that are actually meant by the provisions, its creators, how proximate the creator and their arrangements need to be to the work creation and the right action in case the created works had multiple contributors during the creative process. It is evident that much relies on the interpretations of the court. Notably, Ciani (2019) stated that they were certain that the provisions had no room for AI-generated works due to lack of an author required in the Copyright law, on the other hand, the researchers have created a legal fiction of authorship which means that the copyright is vested upon a party who not the author-in-fact.
One of the legal manners taken includes vesting the copyright in individuals other than the authors. Currently, the common law system has not objected to the attribution of copyright to individuals who are not the authors. For example, Copyright law is allowed to protect the interests of involved professionals apart from the author. For instance in a scenario of producers to phonograms for broadcasters. The Copyright system in the Latin-Germanic is already struggling to deal with this strategy due to the significant structures dependent extensively on the “intellectual” link. Therefore, identifying the authorship for AI systems could encompass disruptions to the traditional idea of authorship. More commentators have indicated that this perspective has been abandoned. Since the Directive 2001/29 was implemented, the European copyright law moved their attention towards protection of producers, investors and those contributing to the work creations economically and financially rather than creatively. The perspective change would align with the information society environment that has both an increasing dissociation with those engaging in the creative efforts and those providing the finances and the minor function played by the author’s personality in the creative process. There is also the “multiplayer model” that gives descriptions to the multiple stakeholders in the process of creating the works through AI. This model also illustrates that the efforts being taken by the traditional Copyright law in identifying one author was becoming insufficient and anachronistic.
Another strategy is the work-for-hire doctrine in bypassing the author-in-fact stipulation. The Copyright law already has legal fictions on granting exclusive rights toi subjects that are not authors of the creative works. One example is the work-for-hire doctrine where the employer of individuals who had the work prepared for them is considered to be the copyright holders due to their economic exploitation of the work made for hiring purposes. The doctrine has already worked in vesting rights of economic exploitation to the publishers of collective works and the products to cinematographic works even iod the author was the editor or the individual responsible for production organization. Presumably, this could be considered a fitting framework in handling the challenge of aI authorship since it is a representation of a present mechanism of directly granting copyright to a legal individual that is not the author-in-fact. In relation to the AI-generated works, its application would occur for a similar cultural reason behind the initial introduction which is holding out the prospect of an economic reward as an incentive to the party that directly invested into the creation of exclusive works.Granting the exclusive rights to programmers and owners of AI systems may be an incentive important in developing the AI industry further.
The neighboring or sui generis right-type of protection could also be granted to the AI generated works. According to Ciani (2019), this rationale is not aligned with that of Copyright law even if it entails granting copyright as rewarding authorship. Nonetheless, authorship is not considered the central element when it comes to the AI-generated works. Rather the right rationale is granting the exclusive rights on the authors to protect the investments. This rationale will illustrate that copyright may not be the best legal framework to be followed in protecting AI-generated works. Notably, other legal tools that could be incorporated in the EU jurisdiction could be the sui generis right for protection of databases and the sui generis rights to favor the producers and broadcasters.
Over Time, the regimes in the EU have worked to protect the distinct types of investments. Nonetheless, Ciani (2019) suggests introducing the neighbouring right-typoe protection of the AI-created works since it would consider the better accounts on the type of creativity that occurs, will be more consistent with the previous policies and regulatory decisions incorporated by member states in this field. This is better compared to adoption of the copyright features to the particular needs of AI. This recommended solution would also be aligned with the Recital 5 Directive 2001/29/EC which encompasses the member states indicating that there are no new concepts in protecting intellectual property that are required. Therefore, the present Copyright laws and any related rights should be adapted and supplemented for responding sufficiently to the economic realities. The new right is to be customized while fully knowing the present and the potential future of AI technologies after a meticulous comprehension of the distinct automation degrees that could characterize the field of computer-generated creativity.
In conclusion, various scholars have looked into the issue of copyright protection in relation to the AI-generated works within the EU. The EU Copyright law and court interpretations have continuously been revised to determine the best way in which the copyright protection rights are to be granted. It is evident that the UK has made major strides in ensuring that the works receive copyright protection. However, these provisions still have their criticisms. Additionally, it is evident that many scholars do agree that a form of copyright protection is required for AI-generated works which will act as an incentive and boost the growth of this respective industry. Nonetheless, the traditional Copyright laws cannot be implemented in these works due to their distinct nature. Therefore, the commentators expect the riot stakeholders to look into incorporating the proper legal tools that would cover a wider scope of works. The neighboring or sui generis right-type is a legal tool that has attracted major recognition as an effective manner. The EU Copyright law could consider the legal tool in granting rights to the AI-generated works.

RESEARCH METHODOLOGY
The research methodology incorporated in this thesis is the legal dogmatic approach. This approach focuses on the interpretation and systematization of the material within the current law. Additionally, the approach also aims at providing substantial justifications to the interpretations of current eU Copyright law. This approach will be effective in answering the two primary research questions of the thesis which include whether the current EU Copyright legislation has adequately canvassed copyr-ght protection for AI-generated works and whether the AI-generated works should actually receive copyright protection and allocated the respective rights. The main objective is to derive appropriate answers after interpretations of the present law and case laws.
The de lege ferenda analysis will be incorporated into the research with a goal of proposing the legislative improvements to the EU Copyright law in ensuring protection for the AI-generated works. To this effect, the analysis will encompass an interpretation of the law and proposes the respective legislative revision. Additionally, the legal history of the Copyright law will also be considered in pursuit of understanding the originality and human author requirement.
Since the main objective of this thesis involves the EU Copyright law, the primary legal sources to be incorporated are the EU Copyright Directives and the case laws handled by the CJEU. furthermore, international treaties and any official papers presented in the EU will be utilized as source material. Additionally, the legal provisions and any case laws among national courts will also be considered from other jurisprudences in a bid to widen the understanding of the research’s topic and examining the resolutions to similar issues incorporated by other jurisprudences. Nonetheless, while the research will undertake some comparative elements, no full comparative analysis is undertaken. This research considers academic articles and other sources as important parts of this thesis as they will provide significant insight to the research questions.

RESEARCH DISCUSSIONS
Research Question One:
Have AI-generated works been adequately canvassed by the European Union Copyright Legislation?”
Overview
This section of the paper focuses on the copyright framework of the EU. This includes having a focused examination of the directives available and how they are being utilized in the EU. The European regulation framework for the copyright stuff and acquis which is concerned with over ten directives. This normally has to harmonize with the essential rights of the broadcasters, producers, performers and ultimately authors. A large percentage of the EU directives which reflect the Member State’s obligations. This is stipulated under the World Trade Organizations commonly known as the “TRIPS” Agreement of 1994. The main aim of the EU Harmonization efforts is focused on enhancing the copyright protection of goods and services which allow the movement of goods freely within the confines of the internal market.
One of the most critical directives revolves on the “Information Society Directive 2001/29/EC” of which it is much concerned with the provision of the common European basis which focuses on the implementation of the common obligations that are set for the WIPO Internet Treaties. As far as WIPO is concerned, comprehension regulation of the things affecting the world, it is important to know that conclusive rights, limitations of catalogue and nonbinding should always be taken into consideration. Another critical directive is the 200/24/EC Software Directive which is more concerned with the computer programs protection. This normally entails the programming of the required software which is critical in the computation world.
Rationale of Copyright law
Copyright as one of the critical instruments it does play a critical role and it does serve two main purposes namely the provision of the sufficient protection of the both the right holders and authors when it comes to the accommodation of the information. And the public. As provided in the “preamble to the so-called WIPO Copyright Treaty (1996)” it is clear that the European Copyright regulation focuses on the maintenance of the balance between the larger public to the rights bestowed to the author based on their education, information access and research. This can be categorized as the preamble 14 in the context of the Infosoc Directive of the 2001.
In the general terms, intellectual property rights are mainly focused on the promotion of and encouragement of technological development. In general terms, the right to exclusively make a critical contribution to the work provision, means that there are a series of financial benefits that need to be understood and appreciated in the context of their work. With less copyright protection, it implies that there is less ride and protection of the development artists. This is an implication that there is a lack of definition when it comes to the future investments in the sense of the direct form. The investment is normally focused on the production of new stuff and legal protection of new things when it comes to the intellectual property when it comes to investment.
Copyright protection Requirement
Copyright protection protects the rights of an individual owner of a program against violation by other individuals or companies. Copyright protection covers both artistic and literary work. However, many state laws have changed in recent years to include computer programs as patentable. The European Union and other regulatory bodies do not issue copyright protection as well as patents to mathematical formulas, scientific ideas, and theories as well as medical discoveries. Intellectual property is however given copyright protection upon proof of the ownership.
The European Union requires individual member states to protect trademarks and copyrights through computer software. This is to avoid violation of copyrights which leads to financial and property losses. The European community law, on the other hand, acknowledges the owner as the controller of the property. The copyright owner is required to reserve the right to reproduce, translate as well as to conduct a public performanc. The copyright holder is required to prove the ownership of intellectual property beyond a reasonable doubt. The territoriality protection principle requires an intellectual property or a computer program to adhere to both the local as well as international protection laws.
Artistic and literary works are also required to be coherent with international standards. The international standards protect components of a computer program which include, manuals programming language, as well as the operating system(OS). The copyright protections reserves the right to the owner to prevent use, transfer, or reproduction of work without their consent. Copyright protection is outlined in the Maastricht Agreement 10, under the council directive of 1991, which defines copyright of computer programs as well as their respective legal protection. To achieve its goal, the European members are advocating for a uniform jurisdiction in copyright and software protection to allow that a standard copyright protection requirement all across the European Union members and beyond.
Artificial intelligence has undergone various changes since its inception in the 1950s. Artificial intelligence is intellectual property that should be patented to avoid duplication or any other violation of the individual rights of the owner. Artificial intelligence runs on a computer program and recently the software and other computer programs have been accepted as patentable intellectual property by the European Union and other regulatory bodies. Artificial intelligence, on the other hand, involves a computer program where the programmer is able to control the machine through the output. The European Union has made various adjustments to accommodate the changes in technology particularly the competitiveness brought about by the rise of technology. The European Union has been the leading body in issuing copyrights as well as software development. Artificial intelligence involves the creation of a machine with human characteristics since as perception of the environment and language use Artificial intelligence is amongst the new patentable technological advancement.
The European Union has been at the forefront in giving copyright protection in the member’s state particularly software development and another related computer program. The European Union extended its copyright protection to cover artistic and literary work as well as computer programs in 1988. The European Union has been able to respond to the related challenges particularly about the violation of terms and conditions of the copyright. Another common challenge faced by many European countries is piracy which makes European Union score poorly when it comes to copyright protection. However, the European Union(EU) has sought to address the issue of industrial, commercial, and intellectual property under the European community law. European community law defines the rights of the copyright owner which include share or transfer of a program to another person in their method. The European community law, on the other hand, fails to address copyright protection as well as a comprehensive cover of intellectual property.
Artificial intelligence systems can be covered by the Maastricht Agreement 10, under the council directive of 1991, which defines copyright of computer programs as well as their legal protection. The European members are advocating for a uniform jurisdiction in copyright and software protection. The member states have however faced difficulties in implementing copyright and trademark policies, and thus it is becoming equally difficult to enact uniform jurisdiction across the member’s state. The European Union issued a software directive dictating that the owner has the right to prevent any kind of use of his artistic and literary work without their consent. The copyright owner also has extended rights to translation, reproduction, public performance as well as adaptation. The rights of an artificial intelligence programmer may be limited to the exclusive rights as outlined by EU software law. The copyright owner has limited rights in private copying, recompilation as well as backup copies. European Union copyright outlines that an individual should be under renewable ownership of a computer program for fifty years. The owner of a given artificial program, therefore, needs to renew the ownership contract after every 50 years. The Confederation of European Users Association(CECUA) clearly defines the appropriate expression of a given computer program and its protection program by both the state and the international community.
In conclusion, artificial intelligence is eligible for protection by the European Union regulations under the category of computer programs. The programmer and the owner of the copyright reserve the right to copy, translate, adapt as well as to conduct a public performance about the program. European Union should, on the other hand, enact the principle of uniformity in software development and beyond across the member states.
The Originality Requirement
European Union Copyright Directive
The European Union refers to the originality requirement in three primary directives. The three guidelines include the Software Directive, the Database Directive, and the Term Directive. According to all directives above, the originality of the work should prevail to deserve copyright protection. The fact that the three directives are similar in working shows that the EU legislators’ objectives were originality to have similar interpretation to such kinds of works. Nonetheless, it’s arguable that despite the similarity in working, the authors’ intellectual creation has a somewhat stricter definition of the Term Directive. This is because the directives state that the intellectual creation of an author works shows his/her personality. It is worth noting that this clarification is not found in the Software Directive; neither is it found in the Databases Directive. What’s more, the different reasoning fundamental to the directive is in full support of the argument that the originality prerequisite for photographs reflects a firmer test compared to that of both the Software Directive and Database Directive.
Precedent in the CJEU
These sections seek to explore the value of precedent in the CJEU system. According to the EU conventions, a formal system of precedent is nonexistent. Notably, the nonexistence of the precedent was initiated by a specific fear; binding system precedent was not in line with the courts’ purpose. Despite the formula rule of the non-precedent, the CJEU had a steady judgment, and hardly did it differ from the previous jurisprudence. According to different studies, the Data Costa Decision initiated the precedent current system under EU law. Notably, the Data Costa decision established that maturational courts and the CJEU have a multilateral relationship. This means that the ruling of the CJEU has direct effects on every national court, irrespective of the referee courts. Furthermore, the decision of the CILFIT supported the precedent by establishing that the ruling of the CJEU was to be purely authorities in the specific situation; for instance, in a situation where the point of law was similar. Moreover, several situations where the CJEU employed “precedent” term and revised the preliminary decision to approve or disapprove the previous cases. The subsequent
The section explores and analyses CJEU cases on the topic of originality.
Analysis of decisions by the CJEU54
The first case this section seeks to explore is the Infopaq case. Mainly, the Infopaq case explored the Infopaq Company’s obligation to get consent from the appropriate holders of the articles before partly reproducing the materials. The cases’ fundamental was focused on the interpretation of the Invesco Directive, which contends that authors have the right to authorize or ban reproduction of their original “works.” What’s more, the CJEU argued that the interpretation of the author’s works required a reversion to Article 2 of the Berne Convention, from which it could be established that the protection of subject matters assumes that it is similar to being an intellectual Creation.
Additionally, a fascinating aspect of the Infopaq case is that the CJEU offered significant importance to the intellectual act of selecting an arrangement to text snippets. According to CJEU, “Regarding the elements of such works covered by the protection, it should be observed that they consist of words which, considered in isolation, are not as such an intellectual creation of the author who employs them. It is only through the choice, sequence, and combination of those words that the author may originally express his creativity and achieve a result of intellectual creation.” As such, several newspaper articles writers have employed various creative moves that ensure the texts looked original to the extent that one may contend that they were the writers’ intellectual creation.
The second case this section seeks to explore is the Painter Case. While CJEU received several questions in Painers, the essential question was whether a photo fit based on photographs could be published in newspapers, journals, or on the Internet without the rights holder’s consent. Essentially, the court pursued elaboration to the question, is the originality standard for photographs in Article 6 of the Term Directive comprises portrait photographs? In response to this question, the court established that appropriate criterion the valuation of whether a written publication is reflected as the author’s intellectual creation or whether the writer expressed his creative abilities in producing the work by making free and creative choices.
Also, by making those numerous choices, the portrait’s author can mark the created work with his choice mark. Consequently, the court established these creative choices as “In the preparation phase, the photographer can choose the background, the subject’s pose, and the lighting. When taking a portrait photograph, he can choose the framing, the angle of view, and the atmosphere created. Finally, when selecting the snapshot, the photographer may choose from various developing techniques, the one he wishes to adopt or, where appropriate, use computer software.”
The last case this section seeks to explore is the Murphy case. Primarily, the originality standard was additionally illuminated in Murphy’s case. According to the case, the CJEU argued if copyright can be claimed in sporting events per se. In this regard, CJEU responded destructively and explained that an author’s intellectual creation includes that the procedure has to leave room for “creative freedom for copyright.” Ultimately, the court established that because football matches obey the conventions of the game, they have no extra room for such creative freedom; Thus, this can’t be subject to copyright. It is worth noting that the court’s argument shows that work is considered original if it is the result of its author’s creative freedom. Generally, the court decision on the three cases, Infopaq, Painter, and Murphy have clarified and developed the EU concept of originality. Moreover, the idea of “author’s intellectual creation,” which was approved as the criteria for the originality requirement in the Infopaq case entails, according to the mentioned cases, that the author made “free and creative choices” and expressed a “personal touch.” At the same time, the process must leave room for “creative freedom.”
Research Question 2:
Should the AI-generated works be provided with adequate Copyright Protection and be Apportioned the respective rights in the EU?
Notably, this research has acknowledged copyright as the exclusive rights granted to a particular work thus restricting its uses which could be achieved by the protected work. For instance, when one buys a book from the bookstore, fundamentally, one cannot legally duplicate the book. Copyright protection prevents the individual’s capacity to communicate nor share any information from the book. Therefore, in determining whether the AI-generated works need to be protected, a careful consideration of any legitimate jurisdictions should be done prior to granting the protection to these new work types. There should also be a consideration of also the value gained from the protection of AI-created works.
The first thing to do is an Assessment of the justifications provided by the copyright system in permitting or demanding for the copyright protection of AI-generated works. Currently, there are three theoretical justifications that have been provided by the EU Copyright law. The three include the personhood, reward and the incentive theories. In consideration of the personhood theory, persons do have a right of protection to their personhood and their personal autonomy. The creation process will involve an individual occupying a product that one has created as an extension of oneself. For the AI-created works, will not be termed as property that are to be protected by copyright due to AI lacking personhood from which the product is considered an extension the personhood theory is evidently not a proper theory to justify the protection of AI-generated works.
In consideration of the reward theory which also has a relation to the personhood theory, it encompasses the individuals investing time, money and energy in the creation of intellectual products having special financial or individual needs to get the protection due to their investments. The Reward theory does receive support for the labor theory created by John Locke which indicates that persons have natural ownership rights in relation to their bodies. These rights are to be granted depending on the products they have created in the existence of individual labor. Labor is based on the notion that since value is created from value, the individual that has given birth to a work needs to enjoy the respective value. Nonetheless, this theory only considers the labour of human beings as the sole bodies able to make property. Therefore, the AI-created works will not be considered as property needing protection in respect to this theory.
As per the incentive theory, the primary objective of copyright protection is the promotion of creative works in order to benefit the public. Copyright is utilized as a motivational tool to encourage the creative activities of the authors as the latter are granted exclusive control and rights for the works they have generated. It is prudent to state that the AI systems cannot encourage the generation of works and hence the theory is also not applicable. However there are human beings who are behind sione of these AI systems and will require this kind of incentive. Copyright protection is required for promoting the utilization and development of the AI systems. Without copyright protection, the users will lack an incentive for producing any works through AI technologies and the developers will lack the incentive to invest in developing the systems. Therefore, granting copyright protection to the individuals behind the ASI-generated works with human authors is bound to encourage the latter and also companies choosing to invest into the bne technologies and their utilization. According to Hilty and Synodinou who conducted independent research, they indicate that the economic rationale for the investment protection is an applicable justification to have the AI-generated works get legal protection.
In case no protection is granted to the AI-generated works, a negative commercial effect is inevitable. Currently, various companies handling various activities in different fields such as music or gaming are utilizing AI technologies to produce respective works. Furthermore, individual creators are also earning incomes by the help of these AI systems. Therefore, failing to protect the works increases the probabilities of their reproduction by their competitors. This causes a negative economic effect on the original creator of these works. The original creators having this financial incentive in developing the creative machines prompts them to produce software that is aso a resource incentive. Lack of the financial incentive only causes a decline in the society’s valuable creations. Therefore, while the AI-generated works challenge the theoretical justifications provided by EU Copyright law due to their humanistic approach, the incentive theory does justify the protection of the AI-generated works even though it means granting them to other human beings that were behind the AI systems but were not the author-in-fact of the AI systems.
Apart from copyright protection of AI-generated systems creating a significant incentive, there is a need to establish if the society will benefit from the incentive created. Protection is expected to be granted when the AI-generated world demonstrates bringing a value almost equivalent to the human-generated works. In the present digital environment, many of the works generated by AI technologies are indifferent in both value and substance from the human-created works. While the AI-generated works that are presented currently lack substantial human input, to state that they should not be protected is only being counterproductive and a tenuous argument.
Various scholarly researchers presented real life examples to demonstrate how the various AI systems do have positive effects on the economy and society in general. For instance, the healthcare diagnostics and targeted treatment that depends on AI will increase the quality of care given to patients and ultimately the quality of lives. Numerous AI-generated such as music and books will also improve the quality of people that utilize them. Fromer presentes an information theory on copyrights indicating the value of the expressive world to the society is attached to their ability to communicate knowledge whether it is systematic, factual or cultural and also convey enjoyable expression in itself. The modern AI systems are presenting expressive works that actually contribute to society in both ways. Therefore, it is proper to state that the AI-generated works do contain intelligence wual to the human-generated works. These systems are generating artistic works through the collection of input which already contain human tendencies and preferences. For instance, the Project Next Rembrandt which was an algorithm utilizing valuable portraits done by an old master attracted great value from the people. Mozart and Bach are some of the great composers whose compositions have been utilized in producing AI-generated music which is something that the users enjoy. Therefore, it is evident that AI-created works play a satisfactory role in meeting individual’s demand similar to human-created works. Therefore, since AI-generated works are providing individuals with what they are demanding, it is proper to have the world protected by copyright as it will boost their creations and further meet the interests of the society.
Additionally, there are other practical considerations that indicate that the AI-generated works need to receive copyright protection rights granted by Copyright law. This is because the lack of protection will not be in alignment with the historically flexible interpretation of Copyright legislation. It practically counters the tenet of technology neutrality which has been fundamental in the development of Copyright law in the current digital environment. The European Commission has also explicitly acknowledged the need to have legal principles applied and continues regardless of the involved technology. Furthermore, if the AI-generated works are not granted the protection rights while the human-generated works receive them, enforceability becomes an acute issue. This is because it becomes challenging to prove whether a particular work is generated by a human being or the AI-generated technologies. The courts will find it challenging trying to evaluate each work brought before them considering the technical expertise needed and details related to the creative process. The cases end up being lagged behind before finding proper resolutions and also the affiliated expenses become exorbitant.
Additionally, failing to protect the AI-generated works is an inconsistent political decision due to the present conditions. Currently, the EU legislator and the CJEU are favoring the maximization of IP protection specifically for the high technology innovators. The Europe 2020 Initiative would also have the Commission emphasizing on the target to develop the “innovation union” through an improvement of framework conditions and increasing access to finance resources for further research and innovation purposes. These initiatives ensure that the innovative ideas become products and services that boost growth and employment opportunities. The European Commission also stated that they are cooperating with member states in order to maximize the impact of investment at the EU; thus, increasing their global competitiveness. It would be politically inconsistent to have AI-generated works not protected by the European Copyright law.
Notably, the EU legislation needs amendments to guarantee the AI-generated works are copyright protected. Having the works released into the public domain and denying the creators their proper incentive is counterproductive. The EU should therefore provide particular legal regulations that will ensure there is a harmonization in how the AI-generated works are protected. Failure to do so only increases the risk of member states drafting and implementing domestic rules that could in turn damage the internal market’s functioning. Considering the Infosoc Directive 2001/29/EC preamble 6: failure to have harmonization at the higher community level will lead to legislative activities at the domestic levels as they try to respond to the technological problems. Their decisions could result in considerable differences in protecting the AI-generated works which further restrict the free movement of services and products due to intellectual property. Ultimately, the internal market defragments and inconsistencies are experienced within the region’s legislative bodies.
Countries such as the United States and China have initiated legislative processes to ensure the protection of AI-generated works. The UK is already protecting the AI-generated works within its region. These countries do acknowledge the importance of this protection in comparison to having them into the public domain. Therefore, it is time that the EU Copyright law is enhanced to canvass the protection of AI-generated works adequately. This will develop the treason’s overall competitiveness in terms of their economy in comparison to their trading partners. Prior to protecting the works, the region is at risk of losing out to other jurisdiction within the AI domain.
Research Question 3:
Is there a legal tool that needs to be implemented that will ensure granting of copyright protection rights to AI-generated works within the EU?
In acknowledging that AI-generated works need to receive copyright protection, it is only prudent that an effective legal tool is selected that will encompass a broader scope of the works covered. The EU Copyright law is currently not adequately canvassed with the right protection tools that will ensure that all issues are handled. The legal tool selected needs to meticulously handle the ownership predicament when it comes to granting copyright protection to the AI-generated works. Traditionally, the work’s author has always been considered to be its owner. For the AI-generated oworks, the author-in-fact would be considered its owner. However, since AI systems are not identified as legal persons, they have no rights nor responsibilities. Therefore, it becomes challenging to allocate rights to a system needing possible ownership alternatives. However, finding the most justified right holder can be challenging due to the many parties involved. There is no one party that provides considerable human input into the creation process to meet the EU’s originality requirement so that they could claim authorship and ultimately ownership.
The sui generis procedure for intellectual property protection is one that can be adopted by the EU Copyright law in light of providing appropriate protection to the AI-generated works. This procedure entails provision of patent-like protection to the works in entirety and their novel components which will last for a particular period to allow the developers to gain back their costs via a sufficient market lead time. Through this method, the developers get to receive their needed incentives to spend both money and time on other innovative products. It also prevents the monopolization of the scientific base through a few large developers who have the needed financial resources to litigate any potential competition. This sui generis plan is expected to make plans and regulate on the IP rights to be granted to the AI-generated software. Under the plan, the creation of a new agency is created to ensure neutrality and the expertise required in making informed choices especially in the sophisticated field. Various aspects of current IP frameworks are also incorporated into the plan such as copyright, patent and trademark legislation in conjunction with other aspects that have not been covered. The aspects are expected to customize the program to meet the particular needs of the software industry and protect the economic interests of the EU region.
The legislative plan will have some particular elements such as a freeze copyright structure for the AI-generated works which entails those having any current copyright protection rights refiling under the new system. This aims on compiling a comprehensive database of all the outstanding works with their owners. Another element is comprehensive definitions to the works that are to be protected by the legislation. Other elements include the definite period in which the protectionis to be granted and confirmation of its novelty or non-obviousness. This procedure is a very detailed plan with numerous aspects that are to be followed accordingly. Nonetheless, the eeness of its implementation will lead to a situation of AI developers having the much needed incentive to invest time and financial resources to bring in better and innovative AI-generated works.

CONCLUSION
The world is evidently transforming which is apparent in the creative processes. For over seven decades, AI research and the respective technologies have been able to create intelligent neural network algorithms from predictable computer programs. Currently, with the increased sophistication and complexities, AI technologies have been able to produce various creative works with many of them requiring minimal human input. However, these developments face a challenge of copyright protection rights. This is because many countries continue to use the traditional copyright legislation in granting copyright protection rights.as illustrated by the EU Copyright law, the regulations have taken a human approach such that any work protected needs to be affiliated with a human person. Therefore, ana analysis of the current EU legislation clearly demonstrated the lack of adequate canvassing of AI-created works within the respective law. It is understandable that the laws were not created with AI in mind. However, it is high time the stakeholders look into improving the legislation to also encompass the AI-generated works which also require extensive investments and bring about similar benefits as the human-generated works.
Additionally, this research has argued that the EU law needs to grant copyright protection to the AI-generated works in order to boost the region’s competitiveness in relation to that field. The region could look into grafting new regulations that will allow granting of protection to the works while handling the impediments presented by the current regulations. In this case, this research proposed implementing the sui generis rule which will primarily docs on the IP rights of AI-generated works. Its main objective is covering these works thus will be tailored in a manner in which the scope is wide to cover every eligible work while boosting innovation in the field.

Bibliography
‘Artificial Intelligence And Intellectual Property: An Interview With Francis Gurry’ (Wipo.int, 2018) October 2018/No.5 accessed 23 August 2020
accessed 23 August 2020
Andres Guadamuz, ‘Artificial Intelligence and Copyright’, World Intellectual Property Organization [WIPO] Magazine (October 2017), available at https://www.wipo.int/wipo_magazine/en/2017/05/article_0003.html>accessed 23 August 2020.
Assinen S, ‘European Union Copyright Protection For AI-Generated Works’ (Master, University of Turku 2018).
Begoña González Otero and Joao Pedro Quintais, ‘Before The Singularity: Copyright And The Challenges Of Artificial Intelligence’ (Pure.uva.nl, 2018) accessed 23 August 2020.
Bentley, L., and Yin Harn Lee. “Directive 2009/24/EC–on the legal protection of computer programs (Computer Programs Directive).” Concise European Copyright Law (2016): 237-274.
Bently, L., Sherman B., Gangjee, D., and Johnson, P., Intellectual Property Law, Fifth Edition, 2018
Bond Toby and Sarah Blair. ‘Artificial Intelligence & copyright: Section 9 (3) or authorship without an author.’ (2019): 423-423.
Bridy, A., ‘The Evolution of Authorship: Work Made by Code’, Columbia Journal of Law & the Arts, Vol. 39, 2016, available at SSRN: https://ssrn.com/abstract=2836568
Case C-5/08 Infopaq International A/S v Danske Dagblades Forening [2009] ECR I-06569.
Ciaini J, Learning From Monkeys: Authorship Issues Arising From AI Technology (Università degli Studi di Torino, 2019).
Cohen G, ‘Should We Grant AI Moral And Legal Personhood? – New World : Artificial Intelligence’ (New World : Artificial Intelligence, 2016)
Commission, ‘Artificial Intelligence: Commission outlines a European approach to boost investment and set ethical guidelines’. Press release, Brussels, 25 April 2018. Available at , accessed on 23 August 2020.
Commission, ‘Europe 2020: A European Strategy for Smart, Sustainable and Inclusive Growth’ COM/2010/2020 final.
Commission, ‘Strategy, Digital Single Market, Policies, Copyright’, last update 17 July 2018. Available at , accessed on 23 August 2020.
Copyright, Designs and Patents Act 1988.
Durham, A. L., Copyright and Information Theory: Toward an Alternative Model of Authorship, BYU Law Rev., 102 (2004).
E. Rosati, Originality in EU Copyright: Full Harmonization through Case Law (Edward Elgar 2013), at 60.
European Commission, ‘Report On The Protection And Enforcement Of Intellectual Property Rights In Third Countries.Ts’ (2020) accessed 23 August 2020
European Commission, ‘Artificial Intelligence: Commission outlines a European approach to boost investment and set ethical guidelines.’ Press release, Brussels, (25 April 2018). http://europa.eu/rapid/press-release_IP-18-3362_en.htm> accessed on 23 August 2020.
European Parliament, ‘European Parliament Resolution of 16 February 2017 with Recommendations to the Commission on Civil Law Rules on Robotics’ 2015/2103 (INL).
European Union. “Directive 2009/28/EC of the European Parliament and of the Council of 23 April 2009 on the promotion of the use of energy from renewable sources and amending and subsequently repealing Directives 2001/77/EC and 2003/30/EC.” (Official Journal of the European Union 5 2009, 2009).
Ginsburg, J., ‘People not machines: authorship and what it means in the Berne Convention’, International Review of Intellectual Property and Competition Law, Vol. 49. No 2, 2018, pp.131-135.
Ginsburg, J.C., The concept of authorship in comparative copyright law, DePaul L. Rev., 52, 1063 (2003).
Guadamuz, A., ‘Do androids dream of electric copyright? Comparative analysis of originality in artificial intelligence generated works’, Intellectual Property Quarterly, 2, 2017, available at: https://ssrn.com/abstract=2981304
Gunning, D. (2017). Explainable artificial intelligence (xai). Defense Advanced Research Projects Agency (DARPA), nd Web, 2, 2.
Hilty R M. ‘European Copyright, Quo Vadis? Impact of Artificial Intelligence on Copyright Issues: Criteria for Protection’ (2018) Brussels, European Copyright Society, May 25, 2018, 1-7. Available at , accessed on 1 October 2018.
Hristov K. ‘Artificial Intelligence and the Copyright Dilemma’ (2017) 57 The Journal of the Franklin Pierce for Intellectual Property, 431-454.
J. C. Ginsburg, ‘The Concept of Authorship in Comparative Copyright Law’ (2002) 52 DePaul L. Rev. 1063.
Jeanne C Fromer, ‘An Information Theory of Copyright Law’ (2014) 64:71 Emory Law Journal, 71- 128, p. 71.
Lambert P., Computer Generated Works and Copyright: Selfies, Traps, Robots, AI and Machine Learning, European Intellectual Property Review, 39(1), 14 (2017).
Lauber-Rönsberg, A., Hetmank, S., ‘The concept of authorship and inventorship under pressure: Does artificial intelligence shift paradigms?’, Journal of Intellectual Property Law & Practice, Vol. 14, No 7, 2019.
Lukoševičienė Aurelija, „On Author, Copyright and Originality: Does the Unified EU Originality Standard Correspond to the Digital Reality in Wikipedia‟ (2017) 11 Masaryk University Journal of Law and Technology, 215-242. Accessed on 23 August 2020.
Margoni, T. (2016). The harmonisation of EU copyright law: The originality standard. In Mark Perry, editor, Global Governance of Intellectual Property in the 21st Century, pages 85–105. Springer, Switzerland.
Margoni T, Artificial Intelligence, Machine Learning And EU Copyright Law: Who Owns AI? (University of Glasgoq 2018) accessed 24 August 2020
Michaux, B., ‘Singularité technologique, singularité humaine et droit d’auteur’, in Laws, Norms and Freedoms in Cyberspace/Droits, norms et libertés dans le cybermonde, Larcier, 2018, pp.401-416
Perry, M., Margoni, T. , ‘From music tracks to google maps: who owns computer-generated works?’, Computer Law & Security Report, Vol. 26, 2010, pp. 621 – 629.
Phillips J, ‘SUI GENERIS INTELLECTUAL PROPERTY PROTECTION FOR COMPUTER SOFTWARE’ (Cyber.harvard.edu, 1992) accessed 24 August 2020
Pila J & Torremans P. European Intellectual Property Law (Oxford University Press, 2016)
Ramalho, A., ‘Will Robots Rule the (Artistic) World? A Proposed Model for the Legal Status of Creations by Artificial Intelligence Systems’, Journal of Internet Law, Vol. 21, No 1, 2017, pp. 12-25.
Reto M Hilty, ‘European Copyright, Quo Vadis? Impact of Artificial Intelligence on Copyright Issues: Criteria for Protection. Brussels, European Copyright Society’ May 25, 2018. Available at , p. 2.
Schönberger, D., ‘Deep Copyright: Up — And Downstream Questions Related to Artificial Intelligence (AI) and Machine Learning (ML)’, in Droit d’auteur 4.0/Copyright 4.0, Schulthess Editions Romandes, 2018, pp.145-173, available at: https://ssrn.com/abstract=3098315
Spindler G, ‘Copyright Law And Artificial Intelligence’ (2019) 50 IIC – International Review of Intellectual Property and Competition Law accessed 23 August 2020
Stone Peter, Rodney Brooks, Erik Brynjolfsson, Ryan Calo, Oren Etzioni, Greg Hager, et al, ”Artificial Intelligence and Life in 2030”, One Hundred Year Study on Artificial Intelligence: Report of the 2015-2016 Study Panel, Stanford University, 2016.
Synodinou T E. ‘AI Intelligence: Criteria for Protection’ (2018), Available at , accessed on 20 October 2018.
Turing, A. M., ”Computing machinery and intelligence”, Mind Vol. 59, No. 236, Oct., 1950.
Ubertazzi, L.C., I diritti d’autore e connessi. Scritti, II ed., Quaderni di AIDA n.5, Giuffrè, Milano, 21-34 (2003).
V.-L. Benabou, DROIT D’AUTEUR, DROITS VOISINS ET DROIT COMMUNAUTAIRE, Bruylant, Bruxelles, 316 (1997).
Yamamoto Takashi B, ‘AI Created Works and Copyright’ (2018) 48 (1) Patents & Licensing, 1-16, p. 1.
Yanisky-Ravid S. ‘Generating Rembrandt: Artificial Intelligence, Copyright, And Accountability in the 3D Era – The Human-Like Authors Are Already Here – A New Model’ (2017) 659 Mich. St. Law Review (2017), 659-726.

Published by

Essays

View all posts