City Law Tutors

View Original

Impact of Artificial Intelligence on the Current Patent System

Abstract

There are various opinions regarding how regulatory bodies ought to reform to make sure that our current intellectual property system is prepared for such revolutionary developments in AI. To further comprehend the effects this will have on our current system and whether this impact is sufficient to spur legislative reform, the author will first explore the current UK patent system and identify how AI is unable to fit within this framework.

Examples of advancements in AI will be identified and analysed to allow further understanding of how the current patent system is not fit for purpose. Various academic viewpoints on the challenges faced by AI will be discussed and different jurisdictions’ responses examined to remedy these issues, such as expanding the term "inventor".

Suggested remedies will be explored using the doctrine-based legal method to provide an understanding of the various approaches that could be taken to fix the issues earlier identified. The author will predict the future impact and analyse various academic recommendations on how to best handle this situation.

Finally, the author will argue that the adoption of a new form of patent protection specific to AI inventions is the best solution, as it provides for AI innovation whilst allowing the current patent system to remain intact for human inventors.

Chapter 1: Can AI be deemed an inventor?

1.1 Introduction

The purpose of this dissertation is to investigate the impact advances in artificial intelligence (AI) will have on the current patent system. This research will examine the influence of AI developments in depth, drawing on current case law to make predictions for the future of patent law.

To provide greater knowledge to the reader, it will draw on debates and discussions from other academics and approaches adopted by different jurisdictions, as well as critically analyse which approach will provide the best promotion of innovation going forward.

The four main routes highlighted within this work will be: [1]

  • Broadening the definition of an inventor to include humans responsible for AI and thus granting them patent ownership

  • Allowing a patent application to identify AI as an inventor and thereby ownership of the patent

  • Safeguarding AI-devised creations with a completely new type of protection

  • Maintaining the status quo and UK’s current approach to AI devised inventions

The first chapter will provide the reader with an overall understanding of AI and emphasise the present misconceptions and misunderstandings in the legal field about what AI is and its rapid advancements. The chapter will detail the present approach taken to AI-created inventions in the UK, emphasising the ongoing discussion between academics on the negative implications for creativity if the system does not adapt to include AI inventions. The second chapter will critically examine the Thaler v Comptroller-General of Patents, Designs, and Trade Marks case (also referred to as the DABUS case) and scrutinise how the current patent regime deals with AI-created inventions and the implications of this going forward.[2] The author will provide an in-depth understanding of the patent application procedure and on what grounds the UK courts, as well as other jurisdictions, denied the patent submission.[3] These approaches will be analysed to best present and understand the implications and how to move forward.

Chapter three will be used to highlight the main arguments against AI being granted ownership and finally discuss AI in action. The fourth chapter will explore the various academics' viewpoints on this topic such as Foss-Solbrekk and Lauber-Ronsberg who both agree the current regime must change to allow AI innovation to survive.[4] In contrast, Papadopoulou and Guarda disagree and feel innovation is a human attribute only, endorsing AI being refused the title of inventor.[5] The chapter will also explain the primary approaches identified for the future of AI inventions.

Finally, the fifth chapter will conclude this dissertation by recommending to the reader a strategy which is believed to best promote innovation and allow for a fair and just future for AI within the patent sector and the many areas of law that surround it.

1.2 What is Artificial Intelligence?

Historically, AI was first developed in the 1950s and is a discipline of computer science whose main objective is to build non-biological intelligent systems.[6] The advancement of technology to the point that it is capable of human innovation has long been prophesised, with Alan Turing predicting that one day we will have a computer that can learn beyond its original instructions. Turing stated, “it would be a pupil who had learnt much from his master but had added much more by his own work.”[7] Upon reflection, it is safe to say that current AI systems have more than evidenced this prediction.[8]

The distinct lack of clarity over the legal status of AI offered by the current UK legal system is one of the many accusations levelled against it.[9] As the ability of AI is developing at a rapid rate, various stakeholders are concerned about its legal status, underlining the need for this definition to be expanded and standardised.[10] As no unified definition has been adopted it has arguably led to a reluctance for lawmakers to amend current legislation to incorporate AI. Due to this ambiguity and in order to provide clarity when discussing AI-related issues in the legal world, the World Intellectual Property Organisation (WIPO) has resolved that a fundamental definition of AI and the terminology associated with it must be defined.[11]  Furthermore, it is common knowledge that when a core legal concept is not defined in statute, it will be explained in case law. Due to the recent developments and legal issues raised, there has yet to be an opportunity for case law to evolve and provide this clarity, compelling many to advocate for the adoption of a uniform definition.[12]

While no legal definition or terminology of AI has yet been accepted,[13] for the benefit of clarity and understanding of this research, the definition provided by the UK government in its 2017 Industrial Strategy White Paper shall be used.[14] This definition was later adopted by the House of Lords Select Committee on AI in a report from 2018, which it was defined as "technologies with the ability to perform tasks that would otherwise require human intelligence, such as visual perception, speech recognition, and language translation" and provides the seemingly most unified definition.[15]

1.3 What is Intellectual Property?

Intellectual Property (IP) is a widely debated area of law concerning the legal rights of patents, trademarks, copyrights and design rights.[16] In reality, the meaning of IP is set to cover all creative works, inventions and goodwill, allowing legal action to be brought against anyone who attempts to benefit from the work of others.[17]

IP is frequently seen as a critical instrument that provides innovators and creators with the legal capacity to protect and defend their works. [18] Until recently, creativity and ingenuity were entirely human attributes.[19]  This effectively eliminated the potential of AI becoming an inventor, consequently preventing the technology from being permitted to file for a patent. However, since the developments within the sector are advancing rapidly, recent technological breakthroughs allowing AI to invent with little human input, have encouraged and prompted many to express concerns with the ability of our current legislation to govern such advancements.[20] The WIPO encapsulates these concerns stating AI has "the potential to be the most disruptive technology ever." [21]  Academics are calling for immediate assessments of existing legislation before it is too late, citing problems that are not restricted to the sphere of intellectual property law. [22]

1.4 The threats AI pose

As AI evolves, it poses several challenges to the IP protection provided by the current UK patent system,[23] with academics claiming AI innovation has the potential to reshape the nature of innovation itself.[24] The UK has confirmed its belief that AI is making significant contributions to innovation, as evidenced by the receipt of two AI patent applications.[25] Due to these advancements and the quick growth of AI investment, Forbes reports that over $10.5 billion was spent on AI in 2019 alone.[26]

The UK government has recognised the significance and importance of AI, as its potential to enhance productivity and deliver significant economic advantages is clear. The establishment of numerous organisations and institutions, such as the Office for Artificial Intelligence, the Artificial Intelligence Council, and the publication of the UK AI strategy, evidence this further.[27]

The impact of AI on the legal profession has been extensively discussed,[28] with authors such as Gargan and Baker delving deep into the ramifications of such advancements.[29] Academics believe that AI will be the originator of the majority of future inventions, propelling creativity further than ever before,[30] with Papastefanou suggesting AI may introduce a “fourth industrial revolution.”[31] Flett stated, “AI is proving to be an increasingly productive method in generating new inventions and techniques at a pace that is simply not achievable by human creation alone.”[32] It was argued that the use of AI to accelerate the development of several COVID-19 vaccines has fuelled the world's support and conviction in the "public good" that governments anticipate as a result of supporting AI. [33]

The UK Parliament has recognised the potential for AI since the 1980s, when a category of “computer-generated works” was created in section 9(3) of the Copyright Designs and Patents Act 1988.[34] According to the UK government website “unleashing the power of AI is a top priority.”[35] However, the UK is not the only government that intends to capitalise on this development, with the European Commission publishing its digital strategy in 2020, focusing on approaches for AI and publishing a further proposal for harmonised rules in 2022.[36] The approach of supporting AI seems to be one which most of the world nations have adopted, these approaches will be discussed further in chapter two.

1.5 The UK’s current approach

The UK government, however, currently maintains that not all ideas generated by an AI system would be protected by a patent. This was reinforced in the DABUS case in which Dr Stephen Thaler constructed an AI computer named the Device for the Autonomous Bootstrapping of Unified Sentience commonly, dubbed DABUS.[37] Dr Thaler applied for two patents at the UK Intellectual Property Office (IPO), European Patent Office (EPO) and various other jurisdictions for the works DABUS created.[38] The applicant contended in the UK case that he was the applicant by virtue of owning the AI that created the invention.[39]

The application prompted two major concerns. Firstly, is it necessary for the named inventor on a patent application to be human? Secondly, if DABUS cannot itself be named an inventor, what rights if any can be granted to the owner of the AI?[40] It is important to note that typically, a patent and its associated rights rest primarily with the inventor or actual deviser of the invention. Mr Thaler’s argument, that the patent and all rights had been transferred by virtue of him being the owner of DABUS is one this research will discuss in depth in chapters 2 and 3.

For now, it is important to note the current UK position to better understand its perceived drawbacks and strengths. The UK IPO denied the patent applications, contending that naming the machine as an inventor does not meet the requirements of the Patents Act 1977 (PA) and that a person (meaning a natural person and not merely a legal person) must be identified as the ‘inventor’. This prompted Dr Thaler to appeal the decision, claiming that DABUS generated a patentable invention and that as the owner he could file for the patent on the AI’s behalf.[41] He stated it was bad law and immoral to illegitimately take credit for an invention that was not his, thus he felt it was appropriate to list DABUS as the inventor on the patent application.

This ruling cemented the UK’s current approach to allow humans that use AI to devise inventions to be named as the inventor with no acknowledgement made to the AI itself. The judgement indicated that had Dr Thaler named himself as the inventor of the application, his applications would have been highly likely to be successful. [42] Many have called for this approach to be revised to better promote innovation and transparency as the capability of AI increases.[43]

The decision in Thaler has sparked a heated discussion regarding the ruling, with several other jurisdictions such as South Africa and Australia initially adopting contrasting philosophies. AI proponents argue that in order to continue to spur investment and innovation, AI-generated inventions should be patentable.[44] According to Nordeman, "AI generated works should not be precluded from obtaining protection by existing related rights on the basis they are AI generated, and should obtain that protection as long as they meet the required criteria.”[45] This argument proposes that AI should not be prohibited from its right to be deemed an inventor merely because it is AI and not human.[46]

In the UK, section 7 and 13 of the PA requires a “person or persons” to be identified as the inventor in a patent application.[47] This was cited in the initial judgment of DABUS as the main reason why AI cannot be deemed an inventor.[48] Furthermore, Article 81 of the European Patent Convention (EPC) similarly requires a “natural or legal person” as being the only entity to transfer ownership of proprietary rights.[49] The USA also takes this approach with Title 35 of the US Code permitting only “Individuals” as inventors of patents.[50]  In 2018, five regions, which together account for 80% of all patent applications worldwide, mandated an inventor must be a person.[51]

However, whilst most jurisdictions’ legislation agrees, it is important to note that this legislation was devised at a time when it was not foreseen that anything other than a human could become an inventor.[52] Arguably this is an outdated approach and must be reconsidered to allow future innovation. As society adapts and develops the law must in turn move with it otherwise it will not provide sufficient protection to allow investment in AI.[53] By remaining unspoken on the issues AI is posing, the law would be simply sitting by while the world around it changes. Surely if the current system was devised before AI was created and could be taken into consideration, it should be replaced or at least amended now such a breakthrough has occurred. The author fears if change is not made then the future of innovation is based on uncertainties and falsehoods with owners of AI taking credit for work that is not theirs by virtue of ownership.[54]

Chapter 2: A Patent Overview

2.1 What is a Patent?

Patents have always been utilised to increase creativity and allow breakthroughs in human creation by promoting innovation.[55] The UK government has declared that “patents must provide the right incentives to AI development, while continuing to promote human creativity.”[56] To better comprehend the consequences of AI on patent law, it is necessary to first define a patent and its application process.

At its simplest, a patent is a form of intellectual property protection obtained by an inventor against their invention awarded by the IPO.[57] A patent is a negative right that prevents others from utilising the protected invention without prior consent. Arguably it confers two rights: the moral right to be recognised as the inventor and the monetary right to be compensated for the invention.[58] When a patent is obtained, the patented item becomes the property of the inventor and can thus be sold or licenced.[59] Patents do not protect the physical appearance of information, which may be covered by trade marks, copyright, or design rights, instead, patents are used to protect technical innovations.

The incentive for disclosing your invention through a patent application is a monopoly right for a finite duration of up to 20 years of market exclusivity. This empowers the patentee to seek an injunction to prohibit others from infringing on their rights and to seek compensation for any damages caused by such infringement. This also gives the inventor the ability to licence out their invention to others, providing an additional source of revenue.[60]

The PA is the primary piece of legislation in England governing the creation and use of patents.[61] Section 1(1) specifies four requirements that must be met for an invention to be patented.[62] When seeking to grasp the argument offered by various countries against AI being classified as an inventor, it is critical to understand these prerequisites. As IP laws were written at a time when only human inventors were considered, [63] the requirements for a successful patent application are biased towards human inventors and, as a result, are widely seen as inapplicable to AI inventions.[64] The PA stipulates that the invention must be new, involve an inventive step, be capable of industrial application and not fall into an excluded category as defined in sections 1(2), 1(3), or 1(4).[65]

Various arguments have been used to justify refusing AI the title of inventor, such as the invention is not novel if it has been reviewed by the various persons who maintain the AI, or that it is not ethically correct to grant legal rights to something that is not human.[66] However, the UK Government has admitted that AI is making significant contributions to innovation and has released a new national AI strategy.[67] This leads to the primary reason for this investigation, as AI develops the ability of autonomous creation, how will this fit into the current patent system?

2.2 Requirement One – The Invention Must Be Novel

For an invention to be considered "novel," often referred to as "new," it must not be deemed to be in the public domain.[68] An invention is novel if it does not form part of the state of the art,[69] meaning the information must not be available or communicated to any member of the public “who was free in law and equity to use it as he pleased.”[70] Information can be made available not merely via written or spoken description but "by use or in any other method."[71] If an invention is given to someone under a duty of confidentiality, such as if a non-disclosure agreement is signed, or if an employee discloses to a third party while under a responsibility to an employer, it is deemed not to have been disclosed to the public. [72]

AI must be able to learn and adapt in such a way in which a human can.[73] The development of AI came from the copying of already existing biological principles. By copying what makes a human human could it be argued the machine is as well?  The use of a neural network is one of the most popular approaches to machine learning.[74] This approach was inspired by how the human brain functions and in its simplest form is an artificial neural network that allows the AI to develop independently. This is relevant as even the owner of the AI does not understand how the invention was created, satisfying the first requirement of an invention being novel.  It could be argued that even if the method is not applicable; AI and its creator’s relationship functions the same as an employer and an employee and would not be deemed as in the public domain, concluding AI satisfies the first requirement.

2.3 Requirement Two – There Must Be an Inventive Step

The second requirement, an inventive step, prevents individuals from patenting what already lies in the public domain.[75] For obvious reasons if no inventive step has occurred the applicant could have simply duplicated a prior invention. The need for an inventive step is a factual inquiry that entails determining whether the invention is obvious to a person versed in the art and having regard to any matter which forms part of the state of the art at the date of the invention.[76] Lord Justice Oliver outlined the steps to determine if the invention is obvious: identify the inventive concept, view the invention through the eyes of a normally skilled person, identify any differences and ask if those differences would have been obvious to a skilled man if viewed without any knowledge. [77]

When determining who classes as a person skilled in the art Lord Reid states this is “a skilled technician who is well acquainted with workshop technique and has carefully read the relevant literature.”[78] The person skilled in the art need not have any inventiveness, but should simply be able to determine if the innovation is in the public domain.[79] For this step to be satisfied the invention must not have been obvious to a skilled person with knowledge of the relevant literature and technique who has taken into account any related matter that is already in the public domain at the date of which the invention was created.

Carpenter argues that the creation of a machine learning tool is certainly patentable, yet the outputs of this machine are beyond the control of the originators of the algorithm.[80] AI is not yet able to grow in technological power and understand how to find data and progress algorithms independently. This means that for the machine to develop, a human contribution is still required.

However, “a human who simply provides training data to an AI system is unlikely to be considered the author of the resulting work.”[81] Lauber-Ro¨nsberg observes certain AI systems are only “a tool of a human   inventor”[82] with Brkan and Bonnet concluding that algorithms are likely ineligible for patent protection as they aid but do not invent themselves.[83] This argument would be justifiable had the machine learning algorithm only been used to automatically classify data. However, AI does not just have the application of classifying data as it detects patterns and learns from the data given to allow it to have the ability to create data which had not previously been input.[84]

Furthermore, AI machines are currently able to create unpredictable and new results independent of the programmer’s instructions, collect data and communicate them to the outside world and learn and improve their system, thereby satisfying the requirement of an inventive step.[85] Additionally it is believed that soon AI will be inventing in ways that make it impossible to identify human input, some even argue this is happening already.[86] This is further examined in the next chapter.

2.4 Requirement Three – The Invention to be Capable of Industrial Application

The third requirement is for the invention to be capable of industrial application. Section 4(1) of the PA provides that an invention shall be deemed capable of industrial application if it can be used in any form of industry.[87] This is a simple requirement which is almost always satisfied as someone would not seek to patent an invention if it if it had no industrial applicability. There is no reason to protect an invention if others do not seek to copy it.

2.5 Requirement Four – The Invention Does Not Fall into an Excluded Category

This leads to the final requirement that an invention must not fall within any of the categories excluded by sections 1(2) and 1(3) of the PA.[88] This provision intends to protect the public from those who would seek to act with malice by patenting inventions which could have a significant positive impact for humanity.[89] Discoveries and scientific theories of mathematical procedures are among the excluded categories. [90] Nevertheless, discoveries or ideas with a technical element or make a technical contribution are not covered in this exemption. [91] This in itself is widely debated as various companies have started patenting genetic material, claiming a technical contribution has been satisfied. [92] Methods for executing mental activities, [93]  games, [94] information presentation, and in some situations computer programmes are also not patentable. [95] However, as proven in Gale's Application 1991 and IBM's Application 1999, courts will not refuse to patent an invention just because it is implemented in the form of a computer programme. [96] Thereby widening the scope to allow AI creations to be patentable.

2.6 International Approaches

China has the second largest number of patent filings concerning AI inventions in the world and perceives itself to be “the world’s leader in AI.”[97] It is also in agreement with the author that by providing more legal certainty and protection to AI-related inventions more investment will be secured.[98] Consequently, the National Intellectual Property Administration of China changed its guidelines to reconsider the provisions preventing AI from being granted patents.[99] This allowed for the patenting of business methods improved by AI that could increase the efficiency of business activities.[100]

Furthermore, it provided a distinction between AI-implemented inventions and pure algorithms, clarifying that an algorithm itself is unpatentable, but inventions with implemented algorithms might be.[101] As AI-related inventions are usually part of an implementation this provides evidence that China is moving towards an approach to allow AI inventions to be patented.

South Africa is in agreement with this approach,  being the first country in history to grant AI a patent and as a consequence legally equate it to a human being “capable of intellectual originality and deserving the protection of such originality.” [102] Foss-Solbrekk calls for an “urgent reevaluation of how AI systems should be protected under IP law.”[103] She argues there is a current deadlock with the law deeming AI systems too technical and falling short of constituting creative subject-matter under copyright but not always technical enough to be eligible for a patent, it seems the current system is not built for such a development.[104] Under French law computer algorithms shall not be granted patents leading applicants to avoid mentioning the use of AI in their submission and arguably preventing true transparency.[105] This could have long-term consequences and masks the true influence AI is having preventing understanding of its capabilities. The German government recognises this by ruling whist an inventor must be a natural person, AI can be additionally named.[106]

The decisions of various jurisdictions such as the US, UK, Australia, South Africa, and the EU have been highlighted in the case of DABUS.[107] The argument of a “natural person” was used by most jurisdictions to deny the application with only South Africa and initially Australia granting the patent.[108] Australia later reversed its ruling in line with the comments made by the EPO when referring to its decision that an inventor is a natural person, it seems this has now become the international standard.[109] As most jurisdictions are in agreement with their approach it seems “the relevant European and US provisions of statutory law and case law do not leave any room for the protection of the outputs of AI machines.”[110]

The WIPO has recently held multiple debates on the consequences of AI breakthroughs in the field of intellectual property law. This has been recognised as the first stage in starting the conversation for global legislative change.[111] There are various legal challenges in the classification and protection of AI systems.[112] Both national and international law have taken the approach that it is necessary for a natural person to be named in a patent application.[113] Common law countries such as India, the UK and New Zealand have drafted additional provisions to protect “computer-generated” works showing universal agreement that AI will reshape the law.[114] Singapore has been utilising AI to find solutions to its ageing population for years, the country has chosen a trust-based approach for fear that premature legislation might stifle innovation.[115] In 2019 it proposed a new AI governance model to provide a set of consistent definitions and principles to allow for greater certainty and promote the adoption and trust of AI.[116] This evidences the author’s argument that further guidance is needed to allow protection for and against AI.[117]

Kitchin discusses in depth the widespread use of intelligent systems and the requirements for the law to develop a further understanding of these systems to be able to best regulate their use.[118] With academics going as far as to state that digital literacy should be taught as part of the teaching of general legal skills.[119] How can we seek to regulate what we do not understand? It is a simple concept for a not so simple topic. The author will not attempt to fully explain in depth how AI works or how AI has developed over time, instead they aim to show how such advances must be acknowledged by the legal system.

Chapter 3: Developments of the law

3.1 The Patent Application Process

Having examined the requirements for a successful patent it is hard to see how AI inventions cannot satisfy these four necessities. However, when examining the patent application process in depth it is easy to see where AI fails.

When applying for a patent in the UK it is important to remember that there are two available types. The first is a national patent, which is issued by the UK Patent Office in accordance with the PA.[120] The second type of patent is a European patent granted by the EPO under the EPC.[121] This application can only be approved if the patent is first validated in the UK. It should be noted that a patent applicant can also use the Patent Cooperation Treaty to petition for WIPO international coverage.

Under UK law any person can apply for a patent. However, the DABUS decision emphasised that the patent belongs to the inventor unless another person is entitled to own the invention by virtue of any rule of law or any legally enforceable agreement existing at the time the invention was made.[122] This most commonly occurs when the inventor is an employee who invents the item covered by the patent application during the course of his employment. Therefore, causing a grey area of law where some attempt to argue AI is the employee of its owner.[123] This will be analysed in more depth later.

The first step in an application is to complete a Patent From 1. This form comprises of twelve parts, each self-explanatory. However, the main issue arises from part 7 where the applicant must identify the inventor.[124] As previously discussed legislation provides that an inventor must be a human being, with recent case law supporting this requirement.[125] Guarda claims AI can now satisfy all of the four requirements needed for an invention to be patentable, however, states “one of the biggest issues appears to be the absence of a natural person beyond the invention.”[126] The Italian Industrial Property Code states that “the first and last name of the inventor must be mentioned in the patent filing”[127] implying that should this be absent the application would be rejected. Similarly, without the inventor being a named person in a UK application it will be deemed incomplete and unsuccessful. Therefore, as AI cannot be named as the inventor there is no way for it to be granted a patent.[128] The concept of rejecting an application if the inventor is not named as a human is not new with the EPO, the UK Intellectual Property Office and the US Patent Office, already ruling that AI systems or machines cannot be considered the inventors of patentable inventions. [129] 

As discussed in chapter two, it seems that this requirement is universal with numerous jurisdictions providing that an inventor must be human in their legislation.

3.2 Public Concerns

Organisations such as the British civil liberties group Big Brother Watch are spreading misinformation regarding AI and in particular facial recognition, creating further uncertainty and misunderstanding. In July of 2022, the organisation brought a complaint to the Information Commissioners Office regarding various uses of AI technology in retail stores across the UK.[130] Companies such as StoreWatch Limited are using AI to help prevent in-store crime. Again, proof of AI’s ability to be used for the social good many argue for.

However, without clarity and regulation, this grey area of law is being exploited by many to cause fear in the public and by companies with nefarious motives.

Chan highlights the possibility of facial recognition AI causing widespread discrimination, bias and inaccuracies in data.[131] Whilst he expresses his concerns he argues that a blanket ban would be overly precautionary without fully considering the wide range of uses and benefits.[132] Instead, he suggests promoting its acceptance by developing the technology “in a calibrated fashion taking into account the relative risks and benefits, risk mitigation measures and safeguards based on legal and ethical considerations.”[133] The author agrees with his approach and argues this should apply to AIs ability to be granted intellectual property rights (IPR) claiming this would prevent damage caused by those who seek to use AI for negative reasons.

The recent development of ChatGPT, an AI based chatbot developed by OpenAI, has spurred further discussion regarding AI and its capabilities.[134] The bot is able to learn and generate a response to various questions without instructions on what the correct answer is. The user simply inputs a question such as “What is AI” and the bot generates the response of “Artificial intelligence is the simulation of human intelligence in machines that are programmed to think and learn like humans.”[135]

Recently the AI has passed a final exam for an MBA programme in Pennsylvania.[136] This again shows its capabilities and has also led to widespread fear from those who believe if left unmanaged and unregulated the possibilities for its uses to be exploited are endless. However, LeCun, Meta’s chief AI scientist stated, “in terms of underlying techniques, ChatGPT is not particularly innovative.” The reason it has caused such commotion is this that is the first time this type of AI has been openly released to the public. [137] Questions regarding ownership of ChatGPT’s outputs have also arisen from a copyright perspective with “policymakers sticking to human creativity as the prism through which copyright is granted.”[138] This brings into question current AI capabilities and the protection granted by the current legal system. With AI’s potential contribution to the world economy estimated to be up to $15.7 trillion by 2030 it is easy to see why companies wish to promote its protection but also exploit its lack of regulation.[139]

The author argues for legislative change to allow AI-devised inventions to be patentable without granting AI rights such as citizenship.[140] This will allow for such inventions to be regulated and ease concerns being raised by the public due to the recent spotlight on AI and its many advancements.

3.3 Ethical Concerns

The most convincing argument given by academics is that the moral right given to an inventor by the granting of a patent is a right that can only be given to a human.[141] The debate around the morality of AI and its implications can be argued to be a large contributing factor as to why courts are refusing to see AI as an inventor. Papadopoulou argues an intellectual creation can only derive from the human mind,[142] using the United States district court in the Monkey Selfie case to illustrate his point, in which the court considered the creativity of an animal, not AI, by questioning whether a monkey could be assigned IPRs.[143] The court ruled under copyright law it could not be an author thereby strengthening the argument if a living being can’t be then nor can AI.

Various ethical issues were raised by the WIPO with the necessity for AI to serve social goods being a theme highlighted throughout all AI-related discussions.[144] The WIPO has considered applying the “FATEN” method in order to ensure a fair approach has been taken.[145] The approach intends to promote both trust and transparency which is crucial and has been highlighted throughout this work as essential for the widespread acceptance of AI. The WIPO aims to protect and ensure data created and used by AI is compliant and safeguards our society.[146] Other stakeholders have also been accessed with it being estimated that “20 percent of the world’s energy consumption is due to technology and largely AI.”[147] These concerns are arguably the conscious reasons as to why jurisdictions are not protecting AI inventions.

Moral considerations must also be given to economic imbalances and AI has been shown to widen the poverty gap. Developed countries have more disposable income and attract more investment enabling them to have the capacity to develop their own AI at a faster rate than less developed jurisdictions. As developing countries do not have this capacity due to focusing their income on areas such as improved healthcare. It has been argued that with AI speeding up innovation, these countries will be left behind as developed countries innovate at a faster rate than ever and ensure they profit from these innovations.

However, AI was used in the development of COVID-19 vaccinations, had an investment in such technology not been made the social impact worldwide would have been huge. Claypoole argues strongly against providing IPRs to AI stating “When will a robot yearn to have its own apartment? When will an AI that invented technology want to re-invest its earnings into better marketing for its product? Until any of these things happen, we should not even be discussing the concept of granting legal rights to AI.”[148] His argument encapsulates the opinion and fears of the majority of AI’s opposition. If we grant AI the ability to be named an inventor, will it then be entitled to various other legal rights and privileges?

Whilst currently sceptics assert even the most advanced of AI cannot yet create new ideas as this is too unique to humans for any machine to simulate, even critics admit that the case of DABUS has shown we are getting close. Taweepon also recognises the impact of granting legal rights to AI, his consideration shows that even if we granted inventorship to AI, who would be able to commercialise the exclusive rights?[149] The questions arising out of granting AI rights is endless with the WIPO stating the “policy questions raised are broad and multidisciplinary.”[150] For example, there is the question of liability, if AI can have rights, it can also infringe them. This would cause various issues, such as vicarious liability and provides again another example of the wide scale impacts a change to our current system will have.

Other authors claim advances in AI will reduce innovation costs, resulting in large volumes of patents with the majority being held by a small dominant group that has the best training data.[151] This would cause small players in the market to struggle. But it seems these critics have paid little consideration to competition law which would surely prevent this. This author maintains that this is a weak argument against AI and the benefits far exceed the possibility of this occurring. Additionally, If AI inventions are unable to be patented it could cause investors to shy away from funding these inventions preventing various developments in multiple industries.[152]

Others advocate the slowing down in the development of inventive machines with research indicating up to 35 per cent of UK total employment will be taken over by computer AI the next two decades.[153] The recent introduction of Amazon GO stores have been used to evidence and fuel sceptics who feel soon even skilled professions such as doctors and lawyers will be directly affected.[154] Removal of such jobs will impact developing countries as tasks that are arguably unskilled will be assigned to AI and there will no longer be a need to outsource such roles. However, the more optimistic view is that more well-paid jobs will be generated to cater for the developments arising from the new technology and will offset job displacement.[155] McCorquodale argues the disruption even if short-term will deprive people of their right to make a living and support their families and will interfere with the standard of living.[156] However, with advances being used to develop vaccines and medical science, arguably the benefit outweighs the potential risk.

Garza and Schwitzgebel argue certain advanced AI merit moral and legal status and as such is owed moral obligations such as the right to be named inventor of its invention.[157] In contrast, Mowbray argues against granting moral rights to AI, taking the extreme view that this would entitle them to access to electricity, maintenance and repairs like humans are entitled to clean accessible water.[158] Whilst the author is in agreement that awarding moral rights to AI is a step too far, awarding some legal rights would aid in its protection.[159] McCorquodale stresses his fears for the broad spectrum of human rights that developments in AI are threatening.[160] He provides examples of violations caused by AI such as Facebook’s use of users’ biometric data for its tagging feature or the various deaths caused by automated vehicles.[161] However, this author brings into question his argument, whilst the examples provided are all valid it is arguably not relevant as all technology can be used for wrong if its user intends it to.[162]

Continuing to privilege human creativity over its machine equivalent may ultimately lead to policymakers’ worse fear of slowing down investment and development of inventive machines.[163] The author argues the concerns surrounding information not being shared due to this lack of protection are valid and whilst ethical considerations must be considered the implications of a lack of transparency should be prioritised.[164]

Chapter 4: The Options

4.1 The Solution?

With recent developments in AI being publicised and discussed globally, many are challenging current jurisdictions’ approaches. Google has started to fund an AI program that will begin to write local news articles, alongside its AI company Deep Mind software that can generate music by listening to recordings.[165] Whilst in 2016 an AI portrait titled “The Next Rembrant” was released and a novel written by a Japanese AI program reached the second round of a national literary prize.[166] Additionally, Sophia, a humanoid robot, has been granted citizenship by Saudi Arabia, raising more questions about AI and its potential. [167] Foss-Solbrekk has stated, “AI applications are transforming  society, promising to help people make better, more         informed decisions and promote greater productivity, improved efficiency and raise general well-being.”[168]

The question of whether an AI will be capable of ingenuity and creativity without human input is no longer the main question.[169] When this technology exists the IP system “must be appropriate to deliver the benefits to society.”[170] The ability of AI to produce innovations that are subject to IPRs and so merit protection is the only thing that can be said with certainty. [171]

4.2 Sui Generis or Utility Model

The possibility of the use of sui generis rights could be a solution. These rights recognise and incentivise the investment in databases of information and do not require elements of creativity or innovation to be present, thereby avoiding the key argument used against AI being granted ownership rights.[172] However, whilst it could offer some form of protection for the data being input into the AI it does not afford the complete protection needed for AI’s outputs.[173]

Alternatively, the Utility model might shield AI’s outputs. It offers ‘minor inventions’ some security through a system akin to that of a patent, although its standards are less onerous.[174] While the Utility model is not available worldwide, key stakeholders like China and parts of Europe are utilising its ability to prevent others from commercially exploiting their invention without permission.[175] There is no "inventive step" required for it to be granted and the procedure is also speedier and less expensive making it the better choice for many. [176]

However, stakeholders are hesitant to apply for such protection due to the lack of widespread popularity and fear it would prevent the inventions from receiving universal protection. Jurisdictions have also adopted the position that if an application for a patent and Utility Model are made at the same time, the first prevails, preventing many from applying for its protection as it is less than that of a patent with its duration of security being uncertain and varying between seven to ten years. This lack of certainty makes the current Utility Model another poor choice for safeguarding AI innovation.

4.3 Trade Secrets

Academics have argued the current use of trade secrets afforded by the EU Trade Secrets Directive (EUTSD)[177] could be the solution.[178] However, this poses a serious risk. Trade secret laws are not designed to provide IP protection.[179] Aplin states “it is not self-evident that IP includes trade secrets.”[180] This is of course a problematic approach as the confidentiality used to protect these secrets causes illegal practices such as discrimination in algorithmic decision making harder to detect.[181]

Niwan has made claims that “some things that might ordinary be protected by traditional IPRs... are best kept secret.”[182] However, Foss-Solbrekk disagrees stating “trade secrets are not intended to be used as an instrument to protect IP.”[183] The consequences of the lack of clarity in these sectors due to the use of the EUTSD are extremely serious and widespread.[184] The recent US case in which Michael Robinson was sentenced to death based on evidence produced by computer software, highlights the authors concerns regarding the lack of transparency provided by trade secrets.[185] In this case, the defendant was not allowed to challenge or access information in relation to the AI’s development with the judge arguing trade secrets law would take priority.[186]

The fact a defendant could not access this information when even his life depended on it highlights the main fear of the author, that a lack of transparency could prove for a fatal future. Further, another drawback is that with advances in AI increasing at a rate in which humans cannot comprehend it is highly likely AI could be used to reverse engineer another AI’s innovation which within itself is a violation of a trade secret. Consequently, trade secrets must not be used as a future solution and further protection must be provided for companies to move away from its use to ensure AI and its inventions are used for the correct purpose.

4.4 Owner, Developer User

Papadopoulou acknowledges the contributions humans have made to AI inventions and poses an approach to categorise such contributions. He claims that there are three categories of personnel who aid in AI inventions.[187] The AI's owner is the first. According to his thesis, this person qualifies as either a "natural" or "legal person." The second is the developer who built the AI system; frequently, there is an employment link between the owner and developer. The third is the user of the system, who, like the owner, frequently has an employment-type connection and is thus subject to the same laws that govern employees who invent things while performing their job duties. This would not interfere with the current patent requirement that an invention must be novel and not in the public domain.[188] However, the question of which category could attain authorship is a hard one.

The owner, of course, has invested in the creation of the AI and while they do not have direct input into the data that is used by the AI, the current patent system is based on rewarding and promoting innovation.[189] Without the initial investment, the AI would not have been created and would not have been able to create.

While the developer has authored or co-authored the system, if working as a contractor or employee of the owner, that developer does not have an automatic right to the output of the system they have created, just as a manufacturer of a camera does not automatically gain all IPRs for the photos taken using it.[190]

The role of the user of the system can be considered to have the closest relationship to the creative output. This individual sets the goal and controls the result more than the other two categories mentioned. Again, the user is normally an employee of the owner, causing all IPRs to automatically go to the owner. This approach allows for incentives for AI innovation while not giving AI the legal personality so many fear.

The author argues this approach could provide a remedy to many of the current issues but is not the perfect answer as if AI forms an employee type relationship it still leaves areas for confusion such as would it be entitled to employee rights and what would its status be in the employment world.[191]

4.5 Joint Inventor

Chesterman argues that “where, AI is used to assist with development, the inventor could be the first person to observe and understand the invention, but the devisers of the underlying algorithms should arguably be considered as joint inventors.”[192] Joint inventorship is based on the theory that attributing patents to people who are not the inventors is not acknowledging the real inventor, the AI.[193] This argument is similar to that provided in the DABUS case and has been seen as a key issue that must be addressed.[194] The theory would allow for the current UK patent system to remain intact and proposes to allow AI to be named as an inventor as long as a natural person has made a significant conceptual contribution to the invention and is named on the application.[195] However, arguably, inputting data would not satisfy the requirement of a “conceptual contribution” and thereby cause the issue of AI inventorship to continue.

Guadamuz provides another solution, arguing that each application must be decided on a case-by-case basis.[196] She points to the case of Nova Productions in which the Court of Appeal declared a player’s input in a computer game is not artistic in nature as no skill or labour of an artist was contributed.[197] Guarda and Trevisanello agree with both Guadamuz and Davies that some form of legal subjectivity must be introduced.[198] This evidence suggests that perhaps a subjective approach is best suited due to the various applications and ways AI contributes or invents. Whilst they recognise that “identifying computers as legal persons entails outcomes past inventorship” they claim these outcomes are not problematic and the law would require adjustments due to AI’s development in time anyway.[199] They disagree with Chesterman’s concept of joint inventorship and outright disagree with “forcing the system to confer the ownership of IPRs on a person, who did not contribute to the creation or invention” claiming it is “unjustifiable.”[200]

The author feels the option is not viable as due to the rate AI can invent taking a subjective approach would cause courts to be overran with applications and cases.

4.6 IPO: Option One

The IPO has additionally provided various options in the hopes that one will prove the solution to this issue.[201] The first is to make no legal change.[202] This option would continue to prevent AI inventions from being patentable and would allow decisions such as DABUS to remain in force.[203]

Whilst the author disagrees with this approach, they do acknowledge its benefits, such as allowing patent law to remain consistent with various jurisdictions, an approach the Australian courts took when reversing their original decision to allow AI to be named an inventor.[204] Yet change is imminent in all jurisdictions, as shown by China and with pressure mounting, it seems universal change is likely.

4.7 IPO: Option Two

The second option provided by the IPO is to expand the word inventor to include humans responsible for the AI systems that devise the inventions. This would allow international businesses to continue to function as the approach is compliant with other jurisdictions. The IPO suggested that “The law would make clear that the inventor would be the human who made the arrangements necessary for the AI to devise the invention.”[205]

The original Australian decision was justified by holding that an agent is a noun that can be a person or an object, allowing an AI system to fit into its definition.[206] Judge Jonathan Beach claimed that this is simply “recognising reality” and claimed that by not adjusting our laws, we are not furthering our society.[207] However, Stephens argues against this, stating that to grant AI rights would award it personhood and this is a step too far.[208] She claims the problem is not purely legal but also a philosophical one, but is still one that governments must tackle.[209]

The IPO clarified that under this option, no patent naming the AI as the inventor would be successful, affirming the ruling in DABUS.[210] They further commented that if a human qualifies as the inventor, it is up to them if they wish to disclose AI was involved in the description of their invention, arguably providing a solution to those such as Mr. Thaler who claim it is immoral to not credit AI for its invention.[211] However, providing humans with the option to disclose AI’s involvement with the invention arguably does not go far enough to promote transparency.

The ethical concerns this raises have also been noted and the lack of transparency provided by the IPO’s suggested approach feeds into the fear felt by those who hear about AI and its developments.[212] If AI is to be accepted and not feared, it must first be understood, as humans have always been taught to fear the unknown. If AI is not required to be identified as an inventor, then it leads to ambiguity and a further lack of knowledge and understanding regarding its capabilities.[213]

The author determined that out of the four options provided by the IPO, this was the second most viable. However, it still would not be able to wholly fix the issues identified as whilst the capability of AI inventions to be patented under this option is vital, the lack of transparency proves to be fatal.

4.8 IPO: Option Three

Option three attempts to remedy this by allowing patent applications to identify AI as an inventor. Unlike option two it “would always be transparent that a non-human inventor has devised an invention with no requirement to name a human inventor.”[214] This could be achieved by amending the legislation to allow AI to be named as an inventor or removing the requirement to name an inventor altogether if the invention is AI devised.[215] The closely responsible human would own the rights in the first instance and no rights would be conferred onto AI’s to allow them the ability to apply for their own patents. This would incentivise innovation and allow the benefits of AI investment to continue.

However, it is evident this approach will not be universally accepted and limits will apply for businesses applying for patents internationally.[216] If companies are unable to gain international patent protection for AI, this may discourage them from naming AI as an inventor and again undermining the main benefit of this approach, transparency.[217] Further, there are fears that this change would lead to different attributes being credited to a person skilled in the art. A skilled person is considered to understand the terminology and field in which the invention stands.[218] There are questions of should a notional skilled person have access to AI tools. The ramifications of this option are huge and cover various legal areas; for example, amendments will need to be made to allow a challenge if AI has been wrongly named as the inventor.

The author agrees with the concept of this option, however, feels it can be adapted further to offer the best solution.

4.9 What’s the answer?

Whilst the options provided all have beneficial aspects, the author argues that none provide an ultimate solution that would allow for the issues highlighted in this research to be rectified. In the final chapter, the author will argue a solution to allow for the benefits of AI to be reaped without the negative implications discussed in chapter three.

Chapter 5: The Solution

5.1 The Ideal Option

In the previous chapter the author provided a brief analysis of the various options academics and the IPO have provided as a possible solution to the issues faced by AI and the current UK patent system.[219] While each has its own benefits, none can be deemed to be the perfect solution.[220] In this concluding chapter a final solution will be presented in the hopes of providing another view to allow the UK to best adapt to the advances of AI and benefit from its innovations.

5.2 The Author’s Solution

The solution the author would like to advocate is a hybrid of the fourth option presented by the IPO and an adaptation of the Utility Model.[221] This option allows for the protection of AI devised inventions through a new type of protection and would allow inventions that are AI devised to be protected if a human inventor cannot be identified.[222] The differences between a current patent and this new form must be made clear and have identifiable and distinctive differences.

Both systems would co-exist to allow for a balance to be struck between both human and AI inventors. By allowing the existing system to remain in force the legal implications would still be present, but it would minimise the impact on human creators and innovation.[223] Whilst legal adjustments would need to be made, it would allow for the creation of a new system with AI solely in mind and permit the UK to best address issues and scenarios that come as a consequence of such innovation.

The requirements for an AI specific patent should be based on the four existing requirements under the PA discussed in chapter two. However, a stricter test should be applied to the requirement of an inventive step as AI has the ability to invent in ways that are not always “obvious” to humans and would easily meet the current standards placed on this requirement.[224]

The argument of floodgates has been used in all areas of law and has been feared by courts for decades.[225] The rational fear of courts and the legal system being overrun by various claims or applications is a valid argument against allowing AI innovations to be patented. This solution allows for some form of protection, and by re-examining and hardening this requirement, it would prevent an influx in patent applications due to the ease with which AI can invent and ensure the protection is only provided to inventions that benefit society. Additionally, the various issues highlighted in previous chapters such as vicarious liability would not be cause for concern under this method as the legislation would be created in such a way that clarity is provided as to who is liable in each scenario.[226]

The author also suggests removing the test of obviousness, as this would provide faster protection in recognition of the faster turnover.[227] Additionally, the protection period of 20 years could be lessened to 10 to reflect AI’s ability to develop and invent faster than humans. This would also provide those who scrutinise the rewarding of AI and highlight the concern and implications this could have, with some form of comfort knowing AI is not being treated equal to a human. Providing only 10 years of protection would still be enough to incentivise investment whilst also differentiating the creativity of a human from that of a machine.[228]

5.3 Conclusion

Developments in AI have caused great disruption in all legal sectors, with IPRs being arguably the most affected. As various stakeholders have provided differing opinions and options as to how to remedy this issue, the author argues a new specialised system must be created to accommodate the innovation of AI. Legislation to realise such protection is a matter of urgency particularly if the UK intends to be an AI superpower.[229]

With all academics in agreement that the law must update to remain relevant and capable of managing such developments, it seems the question is not whether technological advances in AI will impact the current system but rather how is best for the current system to adapt to accommodate such changes.

The author argues that the adoption of a new form of patent protection specific to AI inventions is the best solution to remedy such issues.

List of Cases

UK Cases

Biogen Inc v Medeva plc [1995] RPC 25

Gale’s Application [1991] RPC 305

Hoechst Marion Roussel Ltd and others v Kirin-Amgen and others [2005] 1 ALL ER 667

Humpherson v Syer (1887) 4 RPC 407

In Lux Traffic Controls Ltd v Pike Signals Ltd and Faronwise Ltd [1993] RPC 107

Pall Corp v Commercial Hydraulics (Bedford) Ltd (No.2) [ 1990] FSR 329

Spartan Steel & Alloys Ltd v Martin & Co (Contractors) Ltd [1973] QB 27

Technograph Printed Circuits Limited v Mills & Rockley Electronics) Limited [1972] RPC 346

Windsurfing International INC V Tabur Marine (GREAT BRITAIN) LTD [1985] RPC 59

US Cases

Autonomy Corp Ltd v Comptroller General of Patents, Trade Marks & Designs [2008] EWHC 146

Case T - 0935/97 Computer program product II/IBM [1999] ECR 116

Fujitsu's Application [1997] EWCA Civ 1174

HTC Europe Co Ltd v Apple Inc [2013] EWCA Civ 451

IGT v The Comptroller-General of Patents [2007] EWHC 1341

Lantana v Comptroller-General of Patents [2014] EWCA Civ 1463

Macrossan’s Patent Application [2006] EWHC 705

Tate & Lyle Technology v Roquette Frères [2009] EWHC 1312

Thaler v The Comptroller-General of Patents, Designs and Trade Marks [2020] EWHC 2412

Townsend ’s Application [2004] EWHC 482

Vicom/Computer-related Invention [1987] 2 EPOR 74

US Cases

Beech Aircraft Corp. v. EDO Corp., 990 F.2d 1237, 1248 (Fed. Cir. 1993)

Bilski v. Kappos 561 U.S. 593, 130 S. Ct. 3218 (2010)

Commonwealth v Robinson, No.CC201307777 (Pa. Ct. C.P. Allegheny City 4 February 2016)

Naruto v. Slater, No. 16-15469 (9th Cir. 2018)

Australia Cases

Thaler v Commissioner of Patents [2021] FCA 879

List of Legislation

UK Legislation

Copyright Designs and Patents Act 1988

Senior Courts Act 1981

The Patents Act 1977

EU Legislation

The European Patent Convention art 81

Trade secrets: Directive 2016/943

Italian Legislation

Legislative Decree No.30 of 10 February 2005 art.185

US Legislation

35 U.S.C. 100 (Sept. 16, 2012)

International Treaties

International Covenant on Economic, Social and Cultural Rights 1966

Bibliography

Primary Sources

EU Commission Documents

Commission (EC), ‘European Commission digital strategy Next generation digital Commission’ C (2022) 4388

Commission (EC), ‘Terminating the Current Status Quo - Is It Finally Judgment Day for Ai-Derived Patents? Regulation of the European Parliament and of the Council Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act) And Amending Certain Union Legislative Acts' [2021] SEC 167

Commission (EC), ‘Communication to The Commission European Commission Digital Strategy: A digitally transformed, user-focused and data-driven Commission’ C (2018) 7118

Commission (EC), ‘Regulation of The European Parliament and of The Council Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act) and amending certain Union Legislative Acts’ SEC (2021) 167

Singapore Commission Documents

Personal Data Protection Commission Singapore, Model Artificial Intelligence Governance Framework (INFOCOMM Media Development Authority 2019)

Select Committee Reports

Artificial Intelligence Committee, AI in the UK: ready, willing and able? (HL 2018-03, 100)

Intellectual Property Office, ‘National AI Strategy’ (22 September 2021) <  https://www.gov.uk/government/publications/national-ai-strategy/national-ai-strategy-html-version > accessed 31 October 2022

National Intellectual Property Administration of China, Guidelines for Patent Examination 2017

Office for Artificial Intelligence, ‘National AI Strategy’ (Version 1.2, E02674508, 2021)

State Council's opinions on Accelerating the Construction of Great Power via Intellectual Property under the New Circumstances (2015)

Secondary Sources

Journal Articles

Abbott R, ‘I Think, Therefore I invent: Creative Computers and the Future of Patent Law’ [2016] 57 Boston College Law Review 1079

Abbott R, ‘The Reasonable Robot: Artificial Intelligence and the Law’ (2021) E.I.P.R. 43(4) 

Abungu C, ‘Democratic Culture and the Development of Artificial Intelligence in the USA and China’ (2021) Chin J Comp Law, 9(1) 

Adde L and Smith J, ‘Patent pending: the law on AI inventorship’ (2021) J.I.P.L.P. 16(2) 

Aplin T, ‘Right to Property and Trade Secrets’ (2015) Research Handbook on Human Rights and Intellectual Property, Edward Elgar 426–27

Armstrong D and Schwartfeger P, ‘Cyber law: framing the future?’ (2021) 171 NLJ 7923 

Autor D, 'Why are There Still So Many Jobs? The History and Future of Workplace Automation' (2015) 29 Journal of Economic Perspectives 3

Baker J, ‘A Legal Research Odyssey: Artificial Intelligence as Disruptor’ [2018] Law Library Journal 110 (1)

Baldus O, ‘A practical guide on how to patent artificial intelligence (AI) inventions and computer programs within the German and European patent system: much ado about little’ E.I.P.R. 2019, 41(12) 

Barazza S, ‘The UK IPO AI and IP consultation: answers in search of (more) questions?’ (2021) J.I.P.L.P. 16(12) 

Blok P, ‘The inventor’s new tool: artificial intelligence - how does it fit in the European patent system?’ (2017) E.I.P.R. 39(2) 

Bonadio E, McDonagh L and Dinev P, ‘Artificial intelligence as inventor: exploring the consequences for patent law’ (2021) I.P.Q. 1 

Bond T, ‘Patents: AI machine cannot be inventor’ (2020) PLC Mag. 31(10) 

Brkan M and Bonnet G, ‘Legal and Technical Feasibility of the GDPR’s Quest for Explanation of Algorithmic Decisions of Black Boxes, White Boxes and Fata Morganas’ (2020) 11 EJRR 18, 43–44

Castets-Renard C, ‘The Intersection Between AI and IP: Conflict or Complementarity?’ (2020) 51(2) IIC 141 

Chan G, ‘Towards a calibrated trust-based approach to the use of facial recognition technology’ Int J Law Info Tech (2021) 29 (4): 305

Chatterjee S and Sreenivasulu N, ‘Artificial intelligence and human rights: a comprehensive study from Indian legal and policy perspective’ (2022) Int. J.L.M. 64(1) 

Chesterman S, ‘Artificial intelligence and the limits of legal personality’ (2020) I.C.L.Q. 69(4) 

Chesterman S, ‘Book Reviews: The Reasonable Robot: Artificial Intelligence and the Law’ (2022) International and Comparative Law Quarterly, 71, 1 

Collins C, Dennehy D, Conboy K and Patrick M, ‘Artificial intelligence in information systems research: A systematic literature review and research agenda’ (2021) 60 ISSN 0268 

Colonna L, ‘Legal and regulatory challenges to utilizing lifelogging technologies for the frail and sick’ Int J Law Info Tech (2019) 27 (1): 50 

Cornthwaite J, ‘Intellectual Property’ [2014] Corporate briefing, 1-4

Cornwell T, Taylor S, Williams E and York K, ‘Advanced Therapies: Innovation, Regulation and Patentability in the UK’ (2021) 18(3) BSLR: 89 

Daria Kim, Josef Drexl, Reto M. Hilty, Peter R. Slowinski, ‘Artificial Intelligence Systems as Inventors?’ (2021) Patent Journal, Vol 54, No 07

Deshpande R and Kamath K, ‘Patentability of inventions created by AI - the DABUS claims from an Indian perspective’ (2020) J.I.P.L.P. 15(11) 

Dornis T, ‘Of “authorless works” and “inventions without inventor” - the muddy waters of “AI autonomy’ in intellectual property doctrine” (2021) E.I.P.R. 43(9) 

Dunne N, ‘Liberalisation and the legal profession in England and Wales’ (2021) The Cambridge Law Journal, 80

Fabris D, ‘From the PHOSITA to the MOSITA: will “secondary considerations” save pharmaceutical patents from artificial intelligence?’ (2020) IIC 51(6) 

Flett E, Wilson J and Coudounaris S, ‘Terminating the current status quo - is it finally judgment day for AI-derived patents?’ (2022) C.T.L.R. 28(3) 

Foss-Solbrekk K, ‘Three routes to protecting AI systems and their algorithms under IP law: The good, the bad and the ugly’ (2021) Intellectual Property Law & Practice, Vol 16

Free R, Brinck C and Davies S, ‘Patenting inventions created using an AI system’ (2020) C.I.P.A.J. 49(7/8)

Frey Cand Osborne M, ‘The future of employment: How susceptible are jobs to computerisation?’ (2017) 114 Technological Forecasting & Social Change 254, 265

Gabison G, ‘Who holds the right to exclude for machine work products?’ (2020) I.P.Q. 1, 20

George A and Tarr J, ‘AI inventors and status challenges: Australia weighs in on AI inventorship’ (2022) E.I.P.R. 44(2) 

Ginsburg J, ‘People Not Machines: Authorship and What It Means in the Berne Convention’ (2018) 49(2) IIC 131 

Ginsburg J, ‘People Not Machines: Authorship and What It Means in the Berne Convention’ (2018) 49(2) IIC 131

Grund M 'Is German Utility Model Protection Available for Biotechnology Inventions?' (2006) 8(3) BSLR: 123

Guarda P and Trevisanello L, ‘Robots as artists, robots as inventors: is the intellectual property rights world ready?’ (2021) E.I.P.R. 43(11) 

Kerrigan C, ‘Personhood and the law: do crypto and digital assets need their own rules?’ (2021) 11 JIBFL 791 

Kitchin R, ‘Thinking critically about and researching algorithms’ (2017) Information, Communication and Society 20(1)

Knutt E, ‘Artificial intelligence to carry out PPE checks’ (2018) The Journal of the Working Environment, 12, 3 (1) 

Koorndijk J, ‘Adapting to innovations in artificial intelligence: AI as mental steps under the EPO’ (2021) E.I.P.R. 43(12) 

Lambert P, ‘A slippery slope with contract no longer regal? Pandemic pressures on intellectual property freedom and the need for clarity as king’ (2020) E.I.P.R. 42(9) 

Lauber-Ronsberg, A and Hetmank S, ‘The Concept of Authorship and Inventorship under Pressure: Does Artificial Intelligence Shift Paradigms?’ (2019) 14 JIPLP 570

Lebedenko S, ‘Russian innovation in the era of patent globalisation’ (2022) IIC 53(2) 

Legner S, ‘Are works of artificial intelligence in need for further protection?’ (2021) E.I.P.R. 43(2) 

Liddell K, ‘Patient data ownership: who owns your health?’ (2021) J Law Biosci, 8 (2)

Linarelli J, ‘Treaty Governance, Intellectual Property and Biodiv’ ELR 6 1 (21)

Mahop M, 'Intellectual Property Protection of New Varieties of Plants in Sub-Saharan Africa: Overview of Existing Regimes (Part 2)' (2014) 14(4) BSLR: 130

Makridakis S, 'The forthcoming Artificial Intelligence (AI) revolution: Its impact on society and firms' (2017) 90 Futures 46, 57

Marfe M and Taylor S, ‘AI as an inventor? Not (yet) in the UK, rules Court of Appeal’ (2021) Comps. & Law, 41-43 

McCorquodale R, ‘Artificial intelligence impacts: a business and human rights approach’ (2021) 1 CL 11-28 

McMahon A, ‘Patents, Governance and Control: Ethics and the Patentability of Novel Beings and Advanced Biotechnologies in Europe’ (2021) Cambridge Quarterly of Healthcare Ethics, 30 

Merchant P, ‘High Court confirms that an Artificial Intelligence cannot be named as an inventor: Stephen L Thaler V the Comptroller- General of Patents, Designs and Trademarks’ (2021) 18(2) BSLR: 63 

Mewies S, ‘2022: what to expect in the world of technology’ (2022) 45 CSR 10, 147 

Miller J, ‘AI DABUS the inventor’ (2021) 171 NLJ 7944 

Miller J, ‘Man vs machine—a judge decides’ (2021) 171 NLJ 7949 

Montagnani M and Appen A, ‘IP and data (ownership) in the new European strategy on data’ (2021) E.I.P.R. 43(3) 

Mowbray M, ‘Moral Status for Malware! The Difficulty of Defining Advanced Artificial Intelligence’ (2021) 30, CQHE 517

Müller B, 'Recent case law from Germany concerning the database right' (2007) 2 CL 60

Neely E, ‘Machines and the moral community’ (2014) Philosophy and Technology 27(1): 107

O’Sullivan E, ‘Artificial intelligence and inventors: artificial intelligence challenges the limits of European patent law in the DABUS applications’ (2021) E.I.P.R. 43(7) 

Oriakhogba D, ‘What if DABUS came to Africa? Visiting AI inventorship and ownership of patent from the Nigerian perspective’ (2021) Bus. L.R. 42(2) 

Oster J, ‘Code is code and law is law—the law of digitalization and the digitalization of law’ Int J Law Info Tech (2021) 29 (2): 101 

Papadopoulou A, ‘Creativity in crisis: are the creations of artificial intelligence worth protecting?’ 12 (2021) JIPITEC 408

Papastefanou S, ‘Smart Grids and Machine Learning in Chinese and Western Intellectual Property Law’ (2021) 52(8) IIC 989 

Patchett-Joyce M, ‘Cyber law: touching the future’ (2021) 171 NLJ 7959 

Pearlman R, ‘Recognizing Artificial Intelligence (AI) As Authors and Inventors Under U.S. Intellectual Property Law’ [2018] 24 Rich J L & Tech

Schovsbo J and Riis T, ‘To Be and Not to Be and IPR – The Protection of Trade Secrets in the EU’ (2019) 41 EIPR 401

Schwitzgebel E, Garza M, ‘A defence of the rights of artificial intelligences’ (2015) Midwest Studies in Philosophy 39:98

Senftleben M and Buijtelaar L, ‘Robot creativity: an incentive-based neighbouring rights approach’ (2020) E.I.P.R. 42(12) 

Smyth C, ‘High Court dismisses appeal against decision that an “AI genome” is not patentable’ (2021) Comps. & Law, 62-63 

Spatola N, ‘Conscious machines: Robot rights’ (2018) Science 359 (637) 400

Stanková E, ‘Human inventorship in European patent law’ (2021) The Cambridge Law Journal, 80 

Stockdale M, ‘Legal advice privilege and artificial legal intelligence: Can robots give privileged legal advice?’ (2019) IJEP 23 4 (422) 

Strath J, ‘Court of Appeal confirms an AI-machine cannot be named as an inventor’ (2022) C.T.L.R. 28(1) 

Strath J, ‘DABUS is not a person: Patents Court “retires” AI as an inventor’ (2021) C.T.L.R. 27(1) 

Takenaka T, ‘Inventorship Standards for Biotechnology Inventions Under the U.S. and Japanese Patent Acts’ (2021) 52(5) IIC 556  

Teitelbaum J, ‘Analogical Legal Reasoning: Theory and Evidence’ (2015) Am Law Econ Rev, 17 (1): 160 

Tessensohn J, ‘May 2021 law reforms enhance resilience of Japan’s intellectual property system’ (2021) E.I.P.R. 43(10) 

Tschirhart C, ‘References to Artificial Intelligence in Canada's Court Cases’ (2020) Legal Information Management, 20, 1 

Uluc I, ‘Cybersecurity in international arbitration: how to ease the anxiety and who should administer the antidote?’ (2022) Int. A.L.R. 25(1) 

Walsh K, ‘Intellectual Property Rights and Access in Crisis’ (2021) 52(4) IIC 379  

Walz A, ‘A Holistic Approach to Developing an Innovation--Friendly and Human-Centric AI Society’ (2017) 48(7) IIC 757 

Wan Y and Liu Y, ‘The Impact of Legal Challenges on the Evolution of Web-based Intelligent Agents’ (2008) JICLT, 3(2) 

Wiele B, ‘The human-machine synergy: boundaries of human authorship in AI-assisted creations’ (2021) E.I.P.R. 43(3) 

Wright S and Scheltz A, 'The rising tide of artificial intelligence and business automation: Developing an ethical framework' (2018) 61 Business Horizons 823, 825

Books

Chesterman S, ‘Book Reviews: The Reasonable Robot: Artificial Intelligence and the Law’ (2022) International and Comparative Law Quarterly, 71, 1

DeCosta F, ‘Intellectual Property Protection for Artificial Intelligence’ (2017) Westlaw Journal Intellectual Property

Garingan D, ‘Artificial Intelligence in Legal Practice: Exploring Theoretical Frameworks for Algorithmic Literacy in the Legal Information Profession’ (2021) Legal Information Management, 21

Horton F, ‘Tech competency: a new standard’ (2017) Law Institute Journal 9 (9)

Koenig A, ‘The algorithms know me, and I know them: using student journals to uncover algorithmic literacy awareness’ (2020) Computers and Composition 58

Encyclopedia

Halsbury's Laws (5th edn, 2020) vol 79

Official Discussions

Department for Business, ‘Industrial Strategy: Building a Britain Fit for the Future’ (27 November 2017) < https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/664563/industrial-strategy-white-paper-web-ready-version.pdf > accessed 26 October 2022

Intellectual Property Office, ‘Artificial Intelligence and Intellectual Property: copyright and patents’ (29 October 2021) < https://www.gov.uk/government/consultations/artificial-intelligence-and-ip-copyright-and-patents/artificial-intelligence-and-intellectual-property-copyright-and-patents > accessed 05 February 2022

Intellectual Property Office, ‘Chapter 2: Request for grant of a patent (Form 1)’ (Formalities Manual , 4 January 2021) <https://www.gov.uk/guidance/formalities-manual-online-version/chapter-2-request-for-grant-of-a-patent-form-1> accessed 20 May 2022

Sekhon J, ‘Regulating after Brexit’ (Written evidence to the European Scrutiny Committee, July 2022) <https://committees.parliament.uk/writtenevidence/110199/pdf/> accessed 26 April 2023

World Intellectual Property Organisation, ‘WIPO Conversation on Intellectual Property (IP) and Artificial Intelligence (AI)’ (27 September 2019) < https://www.wipo.int/meetings/en/details.jsp?meeting_id=51767 > accessed 05 February 2022

Online Articles

AI scientist’ (ZDNET, 23 January 2023) < https://www.zdnet.com/article/chatgpt-is-not-particularly-innovative-and-nothing-revolutionary-says-metas-chief-ai-scientist/> accessed 26 January 2023

Briscoe T, 'Suit: Facebook facial recognition technology violates Illinois privacy laws' (Chicago Tribune, 1 April 2015) <https://www.chicagotribune.com/news/breaking/ct-facebook-facial-recognition-lawsuit-met-story.html> accessed 5 January 2023

Bryce A, ‘Intellectual Property: IP and Artificial Intelligence’ (Law Society of Scotland, 14 June 2021) < https://www.lawscot.org.uk/members/journal/issues/vol-66-issue-06/intellectual-property-ip-and-artificial-intelligence/> accessed 21 February 2022

Carpenter T, ‘If My AI Wrote this Post, Could I Own the Copyright?’ (World of Tomorrow, February 2022) <If My AI Wrote this Post, Could I Own the Copyright? - The Scholarly Kitchen (sspnet.org)> accessed 18/06/2022

Chui M, Manyika J and Miremadi M, 'Where machines could replace humans – and where they can't (yet)' (McKinsey Quarterly, July 2016) <www.mckinsey.com/business-functions/digital-mckinsey/our-insights/where-machines-could-replace-humans-and-where-they-cant-yet> accessed 5 January 2023

Claypoole T, ‘When to Give Legal Rights to AIs? When they can Dream’ (Womble Bond Dickinson, 15 September 2020), < https://s3.amazonaws.com/documents.lexology.com/2bbb60f0-c236-4246-bd6e-9de9fa01eec7.pdf?AWSAccessKeyId=AKIAVYILUYJ754JTDY6T&Expires=1645476017&Signature=L2OWy6d%2BJMv6d8T8RKj1w5jrkmk%3D#page=1 > accessed 21 February 2022

Columbus L, ‘Top 25 AI Start-Ups Who Raised the Most Money in 2019’ (Forbes, December 2019) < https://www.forbes.com/sites/louiscolumbus/2019/12/22/top-25-ai-startups-who-raised-the-most-money-in-2019/?sh=7f4a536f229b https://www.forbes.com/sites/louiscolumbus/2019/12/22/top-25-ai-startups-who-raised-the-most-money-in-2019/?sh=7f4a536f229b> accessed 18 June 2022

Costa E, ‘Inventing the New World Can AIs have intellectual property?’ (Daily Philosophy, 2021) <https://daily-philosophy.com/emanuele-costa-ai-intellectual-property/> accessed 15 June 2021

Croft I, 'Differences between utility models vs patents: which should you choose?' (Harper James,24 August 2021) <https://harperjames.co.uk/article/utility-models-vs-patents-smarter-ip-strategy/> accessed 26 April 2023

Davies C, ‘UK patent and copyright law reform considered for AI age’ (Pinsent Masons Out-law, March 2021) < UK patent and copyright law reform considered for AI age (pinsentmasons.com)> accessed 16 June 2022

Hannon P, 'Researchers say use of artificial intelligence in medicine raises ethical questions' (Stanford Medicine News Center, 14 March 2018) <https://med.stanford.edu/news/all-news/2018/03/researchers-say-use-of-ai-in-medicine-raises-ethical-questions.html> accessed 5 January 2023

Jennings M, ‘The challenges of Artificial Intelligence in the field of IP’ (AA Thornton, March 2020) <https://www.aathornton.com/the-challenges-of-artificial-intelligence-in-ip/> accessed 15 June 2022

Jong R, 'The Retribution-Gap and Responsibility-Loci Related to Robots and Automated Technologies: A Reply to Nyholm' [2019] Science and Engineering Ethics <https://doi.org/10.1007/s11948-019-00120-4> accessed 5 January 2023

Lemercier L, ‘Evolution of IP protection for artificial intelligence in France’ (Dentons, March 2021) < https://www.dentons.com/en/global-content/publications/2021/march/29/evolution-of-ip-protection-for-artificial-intelligence-in-france> accessed 18 June 2022

Lizarralde M, ‘The real role of AI in patent law debates’ Int J Law Info Tech (2022) 30 (1): 23

Medeiros M and Sanft J, ‘Artificial intelligence and intellectual property considerations’ (Financier Worldwide Magazine, January 2018) https://www.financierworldwide.com/artificial-intelligence-and-intellectual-property-considerations#.YqnorLbMJPZ accessed 15 June 2022

Muller C, 'The Impact of AI on Human Rights, Democracy and the Rule of Law' (Council of Europe, 24 June 2020) <https://rm.coe.int/cahai-2020-06-fin-c-muller-the-impact-of-ai-on-human-rights-democracy-/16809ed6da> accessed 5 January 2023

Nirwan P, ‘Trade Secrets: The Hidden IP Right’ WIPO Magazine December 2017 <https://www.wipo.int/wipo_magazine/en/2017/06/arti cle_0006.html> accessed June 2022

Nordemann J, ‘AIPPI: AIPPI: No Copyright Protection for AI Works Without Human Input, but Related Rights Remain’ (Wolters Kluwer Blog, November 2019) < AIPPI: No copyright protection for AI works without human input, but related rights remain - Kluwer Copyright Blog (kluweriplaw.com)> accessed 17 June 2022

Ozcan S, Sekhon J and Ozcan O, ‘ChatGPT: what the law says about who owns the copyright of AI-generated content’ (The Conversation, 17 April 2023) <https://theconversation.com/chatgpt-what-the-law-says-about-who-owns-the-copyright-of-ai-generated-content-200597> accessed 26 April 2023

Pradhan N, ‘Remodelling Your Artificial Intelligence IP Strategy: Top 3 Reasons to Invest in Your IP’ (Foley & Lardner LLP, February 2020) < https://www.foley.com/en/insights/publications/2020/02/remodeling-artificial-intelligence-ip-strategy > accessed 18 June 2022

Press G, ‘Alan Turing Predicts Machine Learning and the Impact of AI on Jobs’ (Forbes, February 2017) https://www.forbes.com/sites/gilpress/2017/02/19/alan-turing-predicts-machine-learning-and-the-impact-of-artificial-intelligence-on-jobs/#72bcc06d1c2b accessed 18 June 2022

Rosen M, ‘AI Invents – But Should it Get Patents, Too?’ (National Academies of Sciences, Engineering and Medicine, August 2021) <https://issues.org/artificial-intelligence-patents-innovation-rosen/> accessed 18 June 2022

Stephens K, ‘Who Owns an AI-generated Invention?’ (Bird & Bird, 01 December 2019) <https://www.twobirds.com/en/news/articles/2019/global/who-owns-an-ai-generated-invention> accessed 05 February 2022

Taweepon S, ‘Challenges of Future Intellectual Property Issues for Artificial Intelligence’ (Lexology, December 2018) < Challenges of Future Intellectual Property Issues for Artificial Intelligence - Lexology> accessed 18/06/2022

Thoet A, 'What Amazon's cashier-free store could mean for millions of workers' (PBS, 7 December 2016) <www.pbs.org/newshour/science/amazons-cashier-free-store-mean-millions-workers> accessed 5 January 2023

Tovey A, 'Ten million jobs at risk from advancing technology' (The Telegraph, 10 November 2014) <https://www.telegraph.co.uk/finance/newsbysector/industry/11219688/Ten-million-jobs-at-risk-from-advancing-technology.html> accessed 5 January 2023

Verweij G, ‘AI anaysis Sizing the prize’ (2017) PwC <https://www.pwc.com/gx/en/issues/analytics/assets/pwc-ai-analysis-sizing-the-prize-report.pdf> accessed 5 January 2023

Yadron D and Tynan D, 'Tesla driver dies in first fatal crash while using autopilot mode' (The Guardian, 1 July 2016) <https://www.theguardian.com/technology/2016/jun/30/tesla-autopilot-death-self-driving-car-elon-musk> accessed 5 January 2023

Newspaper Articles

Grierson J, ‘Facial recognition cameras in UK retail chain challenged by privacy group’ (The Guardian, 26 July 2022)

Guadamuz A, ‘Artificial intelligence and copyright’ WIPO Magazine (WIPO website, 5 October 2017)

Ray T, ’ChatGPT is 'not particularly innovative,' and 'nothing revolutionary', says Meta's chief

Sankaran V, ‘Concerns mount as ChatGPT passes MBA exam given by Wharton professor AI scores somewhere between a B- and B on the exam’ Independent (London, 24 January 2023)

Tonkin S, ‘Southern Co-op supermarket faces legal fight over 'Orwellian' use of Chinese-made facial recognition cameras which are 'adding customers to secret watchlists in bid to stop shop’ The Daily Mail (London, 26 July 2022)

Williams H, ‘I’m a copywriter. I’m pretty sure artificial intelligence is going to take my job’ The Guardian (London, 24 January 2023)

Other Sources

OpenAI, ‘ChatGPT’ (OpenAI, 30 November 2022) <https://chat.openai.com/chat> accessed 26 January 2023

Footnotes

[1] Emma L Flett, Jenny Wilson and Stephen Coudounaris, ‘Terminating the Current Status Quo - Is It Finally Judgment Day for Ai-Derived Patents?’ [2022] CTLR 28(3), 51.

[2] Thaler v Comptroller-General of Patents, Designs and Trade Marks [2020] EWHC 2412.

[3] Christian Tschirhart, ‘References to Artificial Intelligence in Canada's Court Cases’ (2020) Legal Information Management, 20, 1, pp. 39.

[4] Katarina Foss-Solbrekk, ‘Three routes to protecting AI systems and their algorithms under IP law: The good, the bad and the ugly’ (2021) Intellectual Property Law & Practice, Vol 16, No. 3; Anne Lauber-Ronsberg and Sven Hetmank, ‘The Concept of Authorship and Inventorship under Pressure: Does Artificial Intelligence Shift Paradigms?’ (2019) 14 JIPLP 570, 578.

[5] Anthoula Papadopoulou, ‘Creativity in crisis: are the creations of artificial intelligence worth protecting?’ 12 (2021) JIPITEC 408 para 1; Paolo Guarda and Laura Trevisanello, ‘Robots as artists, robots as inventors: is the intellectual property rights world ready?’ E.I.P.R. 2021, 43(11), 745.

[6] World Intellectual Property Organisation, ‘WIPO Conversation on Intellectual Property (IP) and Artificial Intelligence (AI)’ (27 September 2019) < https://www.wipo.int/meetings/en/details.jsp?meeting_id=51767 > accessed 05 February 2022; Intellectual Property Office, ‘Artificial Intelligence and Intellectual Property: copyright and patents’ (29 October 2021) < https://www.gov.uk/government/consultations/artificial-intelligence-and-ip-copyright-and-patents/artificial-intelligence-and-intellectual-property-copyright-and-patents > accessed 05 February 2022.

[7] Gil Press, ‘Alan Turing Predicts Machine Learning and the Impact of AI on Jobs’ (Forbes, February 2017) https://www.forbes.com/sites/gilpress/2017/02/19/alan-turing-predicts-machine-learning-and-the-impact-of-artificial-intelligence-on-jobs/#72bcc06d1c2b accessed 18 June 2022; Russ Pearlman, ‘Recognizing Artificial Intelligence (AI) As Authors and Inventors Under U.S. Intellectual Property Law’ [2018] 24 Rich J L & Tech. no. 2.

[8] Simon Chesterman, ‘Artificial Intelligence and the Limits of Legal Personality’ [2020] ICLQ 69, 819–844; Matthew Kay. ‘Robot Sophia comes of age’ [2018] 168 NLJ 7803, 22.

[9] Liane Colonna, ‘Legal and regulatory challenges to utilizing lifelogging technologies for the frail and sick’ Int J Law Info Tech (2019) 27 (1): 50.

[10] Christopher Collins, Denis Dennehy, Kieran Conboy and Patrick Mikalef, ‘Artificial Intelligence in Information Systems Research: A Systematic Literature Review and Research Agenda’ [2021] JIP 60.

[11] WIPO (n 6).

[12] Rachel Free, Coreena Brinck and Simon Davies, ‘Patenting inventions created using an AI system’ (2020) C.I.P.A.J. 49(7/8), 9-11; Paul Lambert, ‘A slippery slope with contract no longer regal? Pandemic pressures on intellectual property freedom and the need for clarity as king’ (2020) E.I.P.R. 42(9), 535-538; Bram Van Wiele, ‘The human-machine synergy: boundaries of human authorship in AI-assisted creations’ (2021) E.I.P.R. 43(3), 164-171.

[13] Stefan Papastefanou, ‘Smart Grids and Machine Learning in Chinese and Western Intellectual Property Law’ [2021] International Review of Intellectual Property and Competition Law 52(8) IIC 989.

[14]  Department for Business, ‘Industrial Strategy: Building a Britain Fit for the Future’ (27 November 2017) < https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/664563/industrial-strategy-white-paper-web-ready-version.pdf > accessed 26 October 2022.

[15] Artificial Intelligence Committee, AI in the UK: ready, willing and able? (HL 2018-03, 100).

[16] Senior Courts Act 1981 (SCA) s 73(5).

[17] Jonathan Cornthwaite, ‘Intellectual Property’ [2014] Corporate briefing, 1-4.

[18] SCA (n 16) s 72(7).

[19] Papadopoulou (n 5).

[20] Toby Bond, ‘Patents: AI machine cannot be inventor’ [2020] PLC Mag. 31(10), 76-77.

[21] Dominique Garingan, ‘Artificial Intelligence in Legal Practice: Exploring Theoretical Frameworks for Algorithmic Literacy in the Legal Information Profession’ [2021] LIM, 21, 97; Kathleen Liddell, ‘Patient data ownership: who owns your health?’ (2021) J Law Biosci, 8 (2): 1–50 

[22] Lambert (n 12); Chesterman (n 8); Jan Oster, ‘Code is code and law is law—the law of digitalization and the digitalization of law’ Int J Law Info Tech (2021) 29 (2): 101.

[23] Alison Bryce, ‘Intellectual Property: IP and Artificial Intelligence’ (Law Society of Scotland, 14 June 2021) < https://www.lawscot.org.uk/members/journal/issues/vol-66-issue-06/intellectual-property-ip-and-artificial-intelligence/> accessed 21 February 2022; Yun Wan and Yue Liu, ‘The Impact of Legal Challenges on the Evolution of Web-based Intelligent Agents’ (2008) JICLT, 3(2): 112–119.

[24] WIPO (n 6) [16].

[25] IPO (n 6).

[26] Louis Columbus, ‘Top 25 AI Start-Ups Who Raised the Most Money in 2019’ (Forbes, December 2019) < https://www.forbes.com/sites/louiscolumbus/2019/12/22/top-25-ai-startups-who-raised-the-most-money-in-2019/?sh=7f4a536f229b https://www.forbes.com/sites/louiscolumbus/2019/12/22/top-25-ai-startups-who-raised-the-most-money-in-2019/?sh=7f4a536f229b> accessed 18 June 2022; Nikhil Pradhan, ‘Remodelling Your Artificial Intelligence IP Strategy: Top 3 Reasons to Invest in Your IP’ (Foley & Lardner LLP, February 2020) < https://www.foley.com/en/insights/publications/2020/02/remodeling-artificial-intelligence-ip-strategy > accessed 18 June 2022.

[27] Intellectual Property Office, ‘National AI Strategy’ (22 September 2021) <  https://www.gov.uk/government/publications/national-ai-strategy/national-ai-strategy-html-version > accessed 31 October 2022.

[28] Garingan (n 21), 97.

[29] Jamie Baker, ‘A Legal Research Odyssey: Artificial Intelligence as Disruptor’ [2018] Law Library Journal 110(1), 5—30.

[30] Ryan Benjamin Abbott, ‘I Think, Therefore I invent: Creative Computers and the Future of Patent Law’ [2016] 57 Boston College Law Review 1079.

[31] Papastefanou (n 13).

[32] Flett (n 1).

[33] ibid.

[34] Copyright Designs and Patents Act 1988 (CDPA); Katharine Stephens, ‘Who Owns an AI-generated Invention?’ (Bird & Bird, 01 December 2019) <https://www.twobirds.com/en/news/articles/2019/global/who-owns-an-ai-generated-invention> accessed 05 February 2022; Elaine Knutt, ‘Artificial intelligence to carry out PPE checks’ (2018) The Journal of the Working Environment, 12, 3 (1).

[35] IPO (n 6).

[36] Commission (EC), ‘Terminating the Current Status Quo - Is It Finally Judgment Day for Ai-Derived Patents? Regulation of the European Parliament and of the Council Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act) And Amending Certain Union Legislative Acts' [2021] SEC 167; Commission (EC), ‘Communication to the Commission European Commission Digital Strategy: A Digitally Transformed, User-Focused and Data-Driven Commission’ COM (2018) 7118; Commission (EC), ‘European Commission digital strategy Next generation digital Commission’ C (2022) 4388; Thaler (n 2).

[37] Janet Strath, ‘Court of Appeal confirms an AI-machine cannot be named as an inventor’ (2022) C.T.L.R. 28(1), 22-24; Ella O’Sullivan, ‘Artificial intelligence and inventors: artificial intelligence challenges the limits of European patent law in the DABUS applications’ (2021) E.I.P.R. 43(7), 469-474.

[38] Maria Lilla Montagnani and Antonia von Appen, ‘IP and data (ownership) in the new European strategy on data’ (2021) E.I.P.R. 43(3), 156-163; Jeanice Koorndijk, ‘Adapting to innovations in artificial intelligence: AI as mental steps under the EPO’ (2021) E.I.P.R. 43(12), 771-776.

[39] Jan Miller, ‘AI DABUS the Inventor’ [2021] 171 NLJ 7944, 4 (2); Rohan Deshpande and Karan Kamath, ‘Patentability of inventions created by AI - the DABUS claims from an Indian perspective’ (2020) J.I.P.L.P. 15(11), 879-889.

[40] Strath (n 37).

[41] Thaler (n 2); Calum Smyth, ‘High Court dismisses appeal against decision that an “AI genome” is not patentable’ (2021) Comps. & Law, 62-63; Mark Marfe and Sarah Taylor, ‘AI as an inventor? Not (yet) in the UK, rules Court of Appeal’ (2021) Comps. & Law, 41-43.

[42] ibid.

[43] IPO (n 6); Laura Adde and Joel Smith, ‘Patent pending: the law on AI inventorship’ (2021) J.I.P.L.P. 16(2), 97-98.

[44] Flett (n 1); Desmond Osaretin Oriakhogba, ‘What if DABUS came to Africa? Visiting AI inventorship and ownership of patent from the Nigerian perspective’ (2021) Bus. L.R. 42(2), 89-99; Toshiko Takenaka, ‘Inventorship standards for biotechnology inventions under the US and Japanese Patent Acts’ (2021) IIC 52(5), 556-578; John A. Tessensohn, ‘May 2021 law reforms enhance resilience of Japan’s intellectual property system’ (2021) E.I.P.R. 43(10), 682-684; Svitlana Lebedenko, ‘Russian innovation in the era of patent globalisation’ (2022) IIC 53(2), 173-193.

[45] Jan Bernd Nordemann, ‘AIPPI: AIPPI: No Copyright Protection for AI Works Without Human Input, but Related Rights Remain’ (Wolters Kluwer Blog, November 2019) < AIPPI: No copyright protection for AI works without human input, but related rights remain - Kluwer Copyright Blog (kluweriplaw.com)> accessed 17 June 2022.

[46] Pearlman (n 7).

[47] The Patents Act 1977 (PA) s 7 and s 13.

[48] Thaler (n 2).

[49] The European Patent Convention art 81; Eva Stanková, ‘Human inventorship in European patent law’ (2021) The Cambridge Law Journal, 80, pp 338-365.

[50] 35 U.S.C. 100 (Sept. 16, 2012).

[51] IPO (n 6).

[52] WIPO (n 6); IPO (n 6).

[53] Ryan Abbott, ‘The Reasonable Robot: Artificial Intelligence and the Law’ (2021) E.I.P.R. 43(4), 276-277.

[54] Karen Walsh, ‘Intellectual Property Rights and Access in Crisis’ (2021) 52(4) IIC 379.

[55] Maya Medeiros and Jordana Sanft, ‘Artificial intelligence and intellectual property considerations’ (Financier Worldwide Magazine, January 2018) https://www.financierworldwide.com/artificial-intelligence-and-intellectual-property-considerations#.YqnorLbMJPZ accessed 15 June 2022.

[56] IPO (n 6).

[57] CDPA (n 34); Jonathan Cornthwaite, ‘Intellectual Property’ LNB News 02/06/2014 56; Halsbury's Laws (5th edn, 2020) vol 79, para 304.

[58] Guarda (n 5).

[59] ibid.

[60] Sanft (n 55).

[61] PA (n 48) s 1.

[62] ibid s 1(1); Sanft (n 55).

[63] WIPO (n 6).

[64] Guarda (n 5).

[65] PA (n 47) s 1(1), s.1(2), s.1(3) and s.1(4).

[66] Theodore Claypoole, ‘When to Give Legal Rights to AIs? When they can Dream’ (Womble Bond Dickinson, 15 September 2020), < https://s3.amazonaws.com/documents.lexology.com/2bbb60f0-c236-4246-bd6e-9de9fa01eec7.pdf?AWSAccessKeyId=AKIAVYILUYJ754JTDY6T&Expires=1645476017&Signature=L2OWy6d%2BJMv6d8T8RKj1w5jrkmk%3D#page=1 > accessed 21 February 2022; Aisling McMahon, ‘Patents, Governance and Control: Ethics and the Patentability of Novel Beings and Advanced Biotechnologies in Europe’ (2021) Cambridge Quarterly of Healthcare Ethics, 30, 529–542.

[67] Office for Artificial Intelligence, ‘National AI Strategy’ (Version 1.2, E02674508, 2021).

[68] PA (n 47) s 1(1); Humpherson v Syer (1887) 4 RPC 407.

[69] ibid s 2(1); Lantana v Comptroller-General of Patents [2014] EWCA Civ 1463.

[70] Humpherson v Syer (1887) 4 RPC 407 (LJ Bowen); Biogen Inc v Medeva plc [1997] RPC 1.

[71] PA (n 47) s 2(2).

[72] ibid s 2(4); Pall Corp v Commercial Hydraulics (Bedford) Ltd (No.2) [ 1990] FSR 329.

[73] Papastefanou (n 13).

[74] ibid.

[75] PA (n 47) s 1(1)(b).

[76] ibid s 3.

[77] Windsurfing International INC V Tabur Marine (GREAT BRITAIN) LTD [1985] RPC 59 (Oliver LJ) [63].

[78] Technograph Printed Circuits Limited v Mills & Rockley Electronics) Limited [1972] RPC 346 (LJ Reid) [348].

[79] Biogen Inc v Medeva plc [1995] RPC 25.

[80] Todd A Carpenter, ‘If My AI Wrote this Post, Could I Own the Copyright?’ (World of Tomorrow, February 2022) <If My AI Wrote this Post, Could I Own the Copyright? - The Scholarly Kitchen (sspnet.org)> accessed 18/06/2022.

[81] Stephens (n 34).

[82] Lauber-Ronsberg (n 4).

[83] Maja Brkan and Gregory Bonnet, ‘Legal and Technical Feasibility of the GDPR’s Quest for Explanation of Algorithmic Decisions of Black Boxes, White Boxes and Fata Morganas’ (2020) 11 EJRR 18, 43–44.

[84] Papastefanou (n 13).

[85] Guarda (n 5).

[86] IPO (n 6).

[87] PA (n 47) s 4(1).

[88] PA (n 47) s 1(2) and s.1(3).

[89] ibid.

[90] ibid s 1(2)(a); Macrossan’s Patent Application [2006] EWHC 705; Tate & Lyle Technology v Roquette Frères [2009] EWHC 1312.

[91] Fujitsu's Application [1997] EWCA Civ 1174.

[92] PA (n 47) s 1(2); Article 52 – Patentable Inventions - The European Patent Convention' (Epo.org, 2017) http://www.epo.org/law-practice/legal-texts/html/epc/2016/e/ar52.html accessed 17 December 2020; Kirin-Amgen Inc and others v Hoechst Marion Roussel Ltd and others; Hoechst Marion Roussel Ltd and others v Kirin-Amgen and others [2005] 1 ALL ER 667; Bilski v. Kappos - 561 U.S. 593, 130 S. Ct. 3218 (2010).

[93] Halliburton Energy Services Inc’s Applications [2012] RPC 129.

[94] IGT v The Comptroller-General of Patents [2007] EWHC 1341.

[95] PA (n 47) s 1(2)(c) and s 1(2)(d); HTC Europe Co Ltd v Apple Inc [2013] EWCA Civ 451; Townsend ’s Application [2004] EWHC 482; Autonomy Corp Ltd v Comptroller General of Patents, Trade Marks & Designs [2008] EWHC 146; Macrossan’s Patent Application [2006] EWHC 705.

[96] Vicom/Computer-related Invention [1987] 2 EPOR 74; Gale’s Application [1991] RPC 305; Case T - 0935/97 Computer program product II/IBM [1999] ECR 116; PA (n 47) s 1(3)(a) and s 1(3)(b); Hoechst Marion Roussel Ltd and others v Kirin-Amgen and others [2005] 1 All ER 667; In Lux Traffic Controls Ltd v Pike Signals Ltd and Faronwise Ltd [1993] RPC 107.

[97] Papastefanou (n 13).

[98] State Council's opinions on Accelerating the Construction of Great Power via Intellectual Property under the New Circumstances (2015), Section 11, Part 3.

[99] National Intellectual Property Administration of China Guidelines for Patent Examination 2017; Cecil Yongo Abungu, ‘Democratic Culture and the Development of Artificial Intelligence in the USA and China’ (2021) Chin J Comp Law, 9(1): 81–108.

[100] ibid Art. 4.2(2), Chapter 1, Part 2.

[101] National Intellectual Property Administration of China Guidelines (n 99) Sections 1 and 2, Chapter 9, Part 2.

[102] Emanuele Costa, ‘Inventing the New World Can AIs have intellectual property?’ (Daily Philosophy,2021) <https://daily-philosophy.com/emanuele-costa-ai-intellectual-property/> accessed 15 June 2021.

[103] Foss-Solbrekk (n 4).

[104] ibid.

[105] Loic Lemercier, ‘Evolution of IP protection for artificial intelligence in France’ (Dentons, March 2021) < https://www.dentons.com/en/global-content/publications/2021/march/29/evolution-of-ip-protection-for-artificial-intelligence-in-france> accessed 18 June 2022.

[106] BPatG 11W (pat) 5/21.

[107] Thaler (n 2).  

[108] Thaler v Commissioner of Patents [2021] FCA 879; Daria Kim, Josef Drexl, Reto M. Hilty, Peter R. Slowinski, ‘Artificial Intelligence Systems as Inventors?’ Patent Journal, Vol 54, No 07 (28 July 2021), p 255.

[109] Thaler (n 2); Amanda-Jane George and Julie-Anne Tarr, ‘AI inventors and status challenges: Australia weighs in on AI inventorship’ (2022) E.I.P.R. 44(2), 61-64.

[110] Guarda (n 5); Oliver Baldus, ‘A practical guide on how to patent artificial intelligence (AI) inventions and computer programs within the German and European patent system: much ado about little’ E.I.P.R. 2019, 41(12), 750-754.

[111] WIPO (n 6); IPO (n 6).

[112] Papastefanou (n 13).

[113] Papadopoulou (n 5).

[114] Guarda (n 5).

[115] WIPO (n 6).

[116] Personal Data Protection Commission Singapore, Model Artificial Intelligence Governance Framework (INFOCOMM Media Development Authority 2019).

[117] Sarah Legner, ‘Are works of artificial intelligence in need for further protection?’ (2021) E.I.P.R. 43(2), 71-73.

[118] Garingan (n 21) pp. 97.

[119] Fabian Horton, ‘Tech competency: a new standard’ (2017) Law Institute Journal 9(9).

[120] PA (n 47).

[121] Peter Blok, ‘The inventor’s new tool: artificial intelligence - how does it fit in the European patent system?’ (2017) E.I.P.R. 39(2), 69-73.

[122] Thaler (n 2)

[123] Michael M. Rosen, ‘AI Invents – But Should it Get Patents, Too?’ (National Academies of Sciences, Engineering and Medicine, August 2021) <https://issues.org/artificial-intelligence-patents-innovation-rosen/> accessed 18 June 2022.

[124] PA (n 47) s 13(2)(a); Intellectual Property Office, ‘Chapter 2: Request for grant of a patent (Form 1)’ (Formalities Manual , 4 January 2021) <https://www.gov.uk/guidance/formalities-manual-online-version/chapter-2-request-for-grant-of-a-patent-form-1> accessed 20 May 2022.

[125] Thaler (n 2); Halsbury's Laws (5th edn, 2020) vol 79, para 305 and 332.

[126] Guarda (n 5).

[127] Legislative Decree No.30 of 10 February 2005 art.185.

[128] PA (n 47) s 7 and s 13; The European Patent Convention (n 49) Article 81 and 60(2).

[129] Guarda (n 5).

[130] Sam Tonkin, ‘Southern Co-op supermarket faces legal fight over 'Orwellian' use of Chinese-made facial recognition cameras which are 'adding customers to secret watchlists in bid to stop shop’ The Daily Mail (London, 26 July 2022); Jamie Grierson, ‘Facial recognition cameras in UK retail chain challenged by privacy group’ (The Guardian, 26 July 2022).

[131] Gary K. Y. Chan, ‘Towards a calibrated trust-based approach to the use of facial recognition technology’ Int J Law Info Tech (2021) 29 (4): 305.

[132] Frank A. DeCosta III, ‘Intellectual Property Protection for Artificial Intelligence’ (2017) Westlaw Journal Intellectual Property pp.1-4.

[133] Chan (n 131).

[134] Henry Williams, ‘I’m a copywriter. I’m pretty sure artificial intelligence is going to take my job’ The Guardian (London, 24 January 2023).

[135] OpenAI, ‘ChatGPT’ (OpenAI, 30 November 2022) <https://chat.openai.com/chat> accessed 26 January 2023.

[136] Vishwam Sankaran, ‘Concerns mount as ChatGPT passes MBA exam given by Wharton professor AI scores somewhere between a B- and B on the exam’ Independent (London, 24 January 2023).

[137] Tiernan Ray, ’ChatGPT is 'not particularly innovative,' and 'nothing revolutionary', says Meta's chief AI scientist’ (ZDNET, 23 January 2023) < https://www.zdnet.com/article/chatgpt-is-not-particularly-innovative-and-nothing-revolutionary-says-metas-chief-ai-scientist/> accessed 26 January 2023.

[138] Sercan Ozcan, Joe Sekhon and Oleksandra Ozcan, ‘ChatGPT: what the law says about who owns the copyright of AI-generated content’ (The Conversation, 17 April 2023) <https://theconversation.com/chatgpt-what-the-law-says-about-who-owns-the-copyright-of-ai-generated-content-200597> accessed 26 April 2023.

[139] Dr. A and S. Rao and Gerard Verweij, ‘AI anaysis Sizing the prize’ (2017) PwC <https://www.pwc.com/gx/en/issues/analytics/assets/pwc-ai-analysis-sizing-the-prize-report.pdf> accessed 5 January 2023.

[140] Martin Senftleben and Laurens Buijtelaar, ‘Robot creativity: an incentive-based neighbouring rights approach’ (2020) E.I.P.R. 42(12), 797-812.

[141] Guarda (n 5).

[142] Papadopoulou (n 5).

[143] Naruto v. Slater, No. 16-15469 (9th Cir. 2018).

[144] WIPO (n 6).

[145] ibid.

[146] WIPO (n 6).

[147] ibid [54].

[148] Claypoole (n 66).

[149] Suebsiri Taweepon, ‘Challenges of Future Intellectual Property Issues for Artificial Intelligence’ (Lexology, December 2018) < Challenges of Future Intellectual Property Issues for Artificial Intelligence - Lexology> accessed 18/06/2022.

[150] WIPO (n 6).

[151] IPO (n 6)

[152] ibid.

[153] Carl Frey and Michael Osborne, ‘The future of employment: How susceptible are jobs to computerisation?’ (2017) 114 Technological Forecasting & Social Change 254, 265; Michael Chui, James Manyika and Mehdi Miremadi, 'Where machines could replace humans – and where they can't (yet)' (McKinsey Quarterly, July 2016) <www.mckinsey.com/business-functions/digital-mckinsey/our-insights/where-machines-could-replace-humans-and-where-they-cant-yet> accessed 5 January 2023; Alan Tovey, 'Ten million jobs at risk from advancing technology' (The Telegraph, 10 November 2014) <https://www.telegraph.co.uk/finance/newsbysector/industry/11219688/Ten-million-jobs-at-risk-from-advancing-technology.html> accessed 5 January 2023.

[154] Alison Thoet, 'What Amazon's cashier-free store could mean for millions of workers' (PBS, 7 December 2016) <www.pbs.org/newshour/science/amazons-cashier-free-store-mean-millions-workers> accessed 5 January 2023; Spyros Makridakis, 'The forthcoming Artificial Intelligence (AI) revolution: Its impact on society and firms' (2017) 90 Futures 46, 57; Scott Wright and Ainslie Scheltz, 'The rising tide of artificial intelligence and business automation: Developing an ethical framework' (2018) 61 Business Horizons 823, 825.

[155] David Autor, 'Why are There Still So Many Jobs? The History and Future of Workplace Automation' (2015) 29 Journal of Economic Perspectives 3.

[156] International Covenant on Economic, Social and Cultural Rights (16 December 1966) Articles 6, 7, 11.

[157] Schwitzgebel E, Garza M, ‘A defence of the rights of artificial intelligences’ (2015) Midwest Studies in Philosophy 39:98–119.

[158] Erica Neely, ‘Machines and the moral community’ (2014) Philosophy and Technology 27(1): 107; Miranda Mowbray, ‘Moral Status for Malware! The Difficulty of Defining Advanced Artificial Intelligence’ (2021) 30, CQHE 517–528.

[159] Nicolas Spatola, ‘Conscious machines: Robot rights’ (2018) Science 359 (637) 400; Jane C. Ginsburg, ‘People Not Machines: Authorship and What It Means in the Berne Convention’ (2018) 49(2) IIC 131; Niamh Dunne, ‘Liberalisation and the legal profession in England and Wales’ (2021) The Cambridge Law Journal, 80, pp 274–307.

[160] Robert McCorquodale, ‘Artificial intelligence impacts: a business and human rights approach’ (2021) 26 1 CL 11; Sheshadri Chatterjee and N.S. Sreenivasulu, ‘Artificial intelligence and human rights: a comprehensive study from Indian legal and policy perspective’ (2022) Int. J.L.M. 64(1), 110-134; Catelijne Muller, 'The Impact of AI on Human Rights, Democracy and the Rule of Law' (Council of Europe, 24 June 2020) <https://rm.coe.int/cahai-2020-06-fin-c-muller-the-impact-of-ai-on-human-rights-democracy-/16809ed6da> accessed 5 January 2023.

[161] Tony Briscoe, 'Suit: Facebook facial recognition technology violates Illinois privacy laws' (Chicago Tribune, 1 April 2015) <https://www.chicagotribune.com/news/breaking/ct-facebook-facial-recognition-lawsuit-met-story.html> accessed 5 January 2023; Danny Yadron and Dan Tynan, 'Tesla driver dies in first fatal crash while using autopilot mode' (The Guardian, 1 July 2016) <https://www.theguardian.com/technology/2016/jun/30/tesla-autopilot-death-self-driving-car-elon-musk> accessed 5 January 2023; Roos de Jong, 'The Retribution-Gap and Responsibility-Loci Related to Robots and Automated Technologies: A Reply to Nyholm' [2019] Science and Engineering Ethics <https://doi.org/10.1007/s11948-019-00120-4> accessed 5 January 2023; Patricia Hannon, 'Researchers say use of artificial intelligence in medicine raises ethical questions' (Stanford Medicine News Center, 14 March 2018) <https://med.stanford.edu/news/all-news/2018/03/researchers-say-use-of-ai-in-medicine-raises-ethical-questions.html> accessed 5 January 2023.

[162] Garry A. Gabison, ‘Who holds the right to exclude for machine work products?’ (2020) I.P.Q. 1, 20-43.

[163] Miranda Mowbray, ‘Moral Status for Malware! The Difficulty of Defining Advanced Artificial Intelligence’ (2021) 30, CQHE 517–528; Marta Duque Lizarralde, ‘The real role of AI in patent law debates’ Int J Law Info Tech (2022) 30 (1): 23.

[164] Inan Uluc, ‘Cybersecurity in international arbitration: how to ease the anxiety and who should administer the antidote?’ (2022) Int. A.L.R. 25(1), 46-65.

[165] Andres Guadamuz, ‘Artificial intelligence and copyright’ WIPO Magazine (WIPO website, 5 October 2017).

[166] ibid.

[167] Taweepon (n 149).

[168] Foss-Solbrekk (n 4).

[169] Simon Chesterman, ‘Book Reviews: The Reasonable Robot: Artificial Intelligence and the Law’ (2022) International and Comparative Law Quarterly, 71, 1, pp 264–266; Michael Stockdale, ‘Legal advice privilege and artificial legal intelligence: Can robots give privileged legal advice?’ (2019) IJEP 23 4 (422).

[170] IPO (n 6).

[171] Mike Jennings, ‘The challenges of Artificial Intelligence in the field of IP’ (AA Thornton, March 2020) <https://www.aathornton.com/the-challenges-of-artificial-intelligence-in-ip/> accessed 15 June 2022.

[172] Björn Müller, 'Recent case law from Germany concerning the database right' (2007) 2 CL 60–67.

[173] John Linarelli, ‘Treaty Governance, Intellectual Property and Biodiv’ ELR 6 1 (21).

[174] Martin Grund 'Is German Utility Model Protection Available for Biotechnology Inventions?' (2006) 8(3) BSLR: 123.

[175] Iain Croft, 'Differences between utility models vs patents: which should you choose?' (Harper James,24 August 2021) <https://harperjames.co.uk/article/utility-models-vs-patents-smarter-ip-strategy/> accessed 26 April 2023.

[176] Marcelin Tonye Mahop, 'Intellectual Property Protection of New Varieties of Plants in Sub-Saharan Africa: Overview of Existing Regimes (Part 2)' (2014) 14(4) BSLR: 130.

[177] Trade secrets: Directive (EU) 2016/943.

[178] Foss-Solbrekk (n 4) p 250.

[179] Tanya Aplin, ‘Right to Property and Trade Secrets’ (2015) Research Handbook on Human Rights and Intellectual Property, Edward Elgar 426–27.

[180] Tanya Aplin, ‘Right to Property and Trade Secrets’ (2015) Research Handbook on Human Rights and Intellectual Property, Edward Elgar 426–27.

[181] Jens Schovsbo and Thomas Riis, ‘To Be and Not to Be and IPR – The Protection of Trade Secrets in the EU’ (2019) 41 EIPR 401.

[182] Prajwal Nirwan, ‘Trade Secrets: The Hidden IP Right’ WIPO Magazine December 2017 <https://www.wipo.int/wipo_magazine/en/2017/06/arti cle_0006.html> accessed June 2022.

[183] Foss-Solbrekk (n 4) p 257.

[184] Trade secrets: Directive (EU) 2016/943; Tanya Aplin, ‘Right to Property and Trade Secrets’ (2015) Research Handbook on Human Rights and Intellectual Property, Edward Elgar 426–27.

[185] Commonwealth v Robinson, No.CC201307777 (Pa. Ct. C.P. Allegheny City 4 February 2016).

[186] ibid.

[187] Papadopoulou (n 5).

[188] ibid.

[189] Pearlman (n 7).

[190] Papadopoulou (n 5).

[191] Chesterman (n 8).

[192] ibid.

[193] ibid.

[194] Thaler (n 2); Strath (n 37).

[195] ibid; Tim W. Dornis, ‘Of “authorless works” and “inventions without inventor” - the muddy waters of ‘AI autonomy’ in intellectual property doctrine’ (2021) E.I.P.R. 43(9), 570-585.

[196] Andres Guadamuz, ‘Artificial intelligence and copyright’ WIPO Magazine (WIPO website, 5 October 2017).

[197] Nova Productions v Mazooma Games [2007] EWCA Civ 219.

[198] Guarda (n 5).

[199] ibid.

[200] Chesterman (n 170); ibid.

[201] Stefano Barazza, ‘The UK IPO AI and IP consultation: answers in search of (more) questions?’ (2021) J.I.P.L.P. 16(12), 1291-1292.

[202] IPO (n 6).

[203] Thaler (n 2); Jan Miller, ‘Man vs machine—a judge decides’ (2021) 171 NLJ 7949, p4 (1).

[204] IPO (n 6); Thaler FCA 879 (n 108) 126–129.

[205] ibid; Collins (n 10).

[206] Marta Duque Lizarralde, ‘The real role of AI in patent law debates’ Int J Law Info Tech (2022) 30 (1): 23.

[207] Thaler FCA 879 (n 108) 129.

[208] Stephens (n 34); Charles Kerrigan, ‘Personhood and the law: do crypto and digital assets need their own rules?’ (2021) 11 JIBFL 791.

[209] Enrico Bonadio, Luke McDonagh and Plamen Dinev, ‘Artificial intelligence as inventor: exploring the consequences for patent law’ (2021) I.P.Q. 1, 48-66.

[210] Thaler (n 2); Phil Merchant, ‘High Court confirms that an Artificial Intelligence cannot be named as an inventor: Stephen L Thaler V The Comptroller- General of Patents, Designs and Trademarks’ (2021) 18(2) BSLR: 63.

[211] Thaler (n 2); Guarda (n 5).

[212] IPO (n 6).

[213] Commission (EC), ‘Communication to The Commission European Commission Digital Strategy: A digitally transformed, user-focused and data-driven Commission’ C (2018) 7118 final; Commission (EC), ‘Regulation of The European Parliament and of The Council Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act) and amending certain Union Legislative Acts’ SEC (2021) 167 final; Marta Duque Lizarralde, ‘The real role of AI in patent law debates’ Int J Law Info Tech (2022) 30 (1): 23.

[214] IPO (n 6); Thaler FCA 879 (n 108).

[215] Cerys Wyn Davies, ‘UK patent and copyright law reform considered for AI age’ (Pinsent Masons Out-law, March 2021) < UK patent and copyright law reform considered for AI age (pinsentmasons.com)> accessed 16 June 2022.

[216] Michael Patchett-Joyce, ‘Cyber law: touching the future’ (2021) 171 NLJ 7959, p13.

[217] ibid; McCorquodale (n 160); Daniele Fabris, ‘From the PHOSITA to the MOSITA: will “secondary considerations” save pharmaceutical patents from artificial intelligence?’ (2020) IIC 51(6), 685-708.

[218] Jennings (n 171).

[219] Dean Armstrong and Paul Schwartfeger, ‘Cyber law: framing the future?’ (2021) 171 NLJ 7923, p15.

[220] Sally Mewies, ‘2022: what to expect in the world of technology’ (2022) 45 CSR 10, 147.

[221] IPO (n 6).

[222] Tamsin Cornwell, Sarah Taylor, Emily Williams and Kiah York, ‘Advanced Therapies: Innovation, Regulation and Patentability in the UK’ (2021) 18(3) BSLR: 89.

[223] Céline Castets-Renard, ‘The Intersection Between AI and IP: Conflict or Complementarity?’ (2020) 51(2) IIC 141.

[224] ibid.

[225] Spartan Steel & Alloys Ltd v Martin & Co (Contractors) Ltd [1973] QB 27.

[226] Taweepon (n 149).

[227] ibid.

[228] Axel Walz, ‘A Holistic Approach to Developing an Innovation--Friendly and Human-Centric AI Society’ (2017) 48(7) IIC 757.

[229] Joe Sekhon, ‘Regulating after Brexit’ (Written evidence to the European Scrutiny Committee, July 2022) <https://committees.parliament.uk/writtenevidence/110199/pdf/> accessed 26 April 2023.