Scientific/Technological Innovations and Intellectual Property Law

Introduction

In the eighteenth century, humanity invented the first element of mass production known as steam power or the application of steam power. Until then, the only power sources for production were human hands and livestock. With steam power, humanity realized that it could produce many goods with less human labor. This led to a significant expansion in commercial sales and thus legal contracts in countries that quickly adopted steam power. The law needed to be updated. From there, humanity's journey reached production with less reliance on human power or human-based production. Our era is about to change dramatically, as before. This evaluation has been made within the main scope of robotics. Technology is advancing as quickly as possible, and in this context, the fundamentals of intellectual property must change. This assessment evaluates how international property rights will fundamentally change, the reasons for the change, and what awaits us in the near future.

Overview

The Need for Intellectual Property Rights and Beyond

Humanity can imagine and find a way to make life easier. In this context, there have always been inventions, from the wheel to the atomic bomb. However, after accepting capitalism, every creation came with a value. Over the years, with developments, when an invention is created in accordance with the laws, the inventor gains exclusivity in using and selling the invention. Therefore, economic value is provided by granting exclusive rights to inventors. On the other hand, not only inventions but other intellectual properties are also made profitable through legal regulations.

Today, humanity is on the verge of new breakthroughs. Some of these, like artificial intelligence (AI), three-dimensional printers (3D printers), quantum computing systems, and genetically modified organisms, have already entered our lives, and some are on the way, such as androids, cities exclusively for robots, AI inventors, and more. Today, a robot holds citizenship rights in Saudi Arabia. Therefore, the reorganization of terms is necessary and inevitable for every legal concept, especially ownership of property and intellectual property rights.

From Human-Based Inventors to AI-Based Inventors

Intellectual property (IP) is a category of property that includes abstract creations of the human mind. However, in the 21st century, abstract creations may not be entirely invented by humans. It has been mentioned that the inventions above come from AI. In this context, it is possible to explain the importance of exclusive rights, especially under patent law.

The essence of granting exclusivity in patent law aims to protect the created value and encourage new and more initiatives. The world tends to see innovation as a driving force to open new fields in many areas. However, even if you have a patent, it only provides protection for 20 years. Additionally, for some inventors or some other people, this period was not sufficient. Therefore, they started to keep how they invented their inventions secret, and as a result, even the European Patent Office had to adopt the Trade Secrets Directive in 2016. Today, the innovation humanity needs may be in the hands of non-human intelligences.

The issue is much deeper than it appears. From the perspective of patent law of the European Patent Office, it has been stated that to be an inventor, one must be human. However, in the near future, people will be able to connect their minds to computers. Even today, experiments continue. In this case, the European Patent Office must first define what a human is or define what a human and a robot are globally through a United Nations resolution. For example, when a human integrates with a computer, they can create a new telecommunication method. When applying for patent protection, the European Patent Office or other patent authorities will have to examine whether the inventor is human or computer.

Views of Official Authorities on Intellectual Property Rights

Definition of Patent and the Need for Trade Secrets

In Europe, the European Patent Convention (EPC) was adopted in 1973. Before that, similar issues were included in the Paris Convention for the Protection of Industrial Property of 1883 and the Patent Cooperation Treaty. According to Article 52 of the EPC, patents are granted for any invention in all fields of technology. The article also states that the invention must be new, involve an inventive step, and be industrially applicable.

Today, patents are based on granting inventors exclusive rights through laws, but for a limited period. If this exclusivity continued indefinitely, it was thought that innovation could not advance further. Therefore, patents can only provide protection to their inventors for 20 years. We have seen that industries have developed and production methods have become more efficient thanks to such innovation protections, reaching very high levels of goods, and arriving in the 21st century. However, some people began to realize that they could always profit by keeping their inventions secret rather than for 20 years. Hence, the issue of trade secrets arose. Today, many known innovations, like the Uber algorithm or the Google search engine algorithm, are kept secret.

The Need for Change in the Industrial Applicability Requirement in Patent Law

When narrowed down to patent law, it is possible to see that some requirements are quite concrete against new technologies. As seen from the explanation of the EPC, patents are possible when they are industrially applicable. Without this condition, it is impossible to provide exclusive rights even if the invention has other parameters. In 1986, the meaning of the word industry was probably first explained by the decision of the European Patent Office's Board of Appeal, and industry means an activity carried out independently, continuously, and for financial gain. Today, the Examination Guidelines explain the term industry in a more detailed and broader sense.

With disruptive technologies, financial gain can be replaced by something else or greatly affected by them. Probably the requirement of industrial applicability is about to become history because, in this new era, knowledge accumulation and the ability to obtain information faster than others will be new elements of commercial gain. Production will likely become digital and personal because 3D printing is already changing the way we think about the production cycle. It has begun to be thought that a company can easily increase its profits through additive manufacturing via 3D printing.

Future Outlook on the Effects of Disruptive Technologies

The Path to Promoting Technology Development May Depend on Non-Humans

As is known, the aim of granting exclusive rights to inventors is for them to profit from their inventions, which helps development. However, humanity is approaching the end of the road for inventions or discoveries. Even ten years ago, scientists and inventors began to depend on computer assistance. Today, even ordinary people cannot live without their mobile phones or computers. Our daily lives have changed, and people have begun to ask many things to the interfaces of their phones.

The Patent Issue on Robotics and the Trade Secret Option

From a different perspective, the protection of intellectual property rights is not always profitable. Profitable goods are important in the role of intellectual property rights. However, this is actually the results of research and development (R&D) investments, the fundamental organization of innovation. While innovators conduct R&D processes, they want to keep their progress secret; otherwise, the R&D work may not be profitable, and this may affect R&D investments due to the lack of economy as a cycle. At this point, trade secrets have once again emerged.

Keeping innovative information secret is very beneficial, especially in several European countries where robotic inventions are significant and valuable. In particular, in some scenarios where reverse engineering is possible, a patent for the legal process may be preferred instead of trying to protect the production process. Moreover, as mentioned above, trade secret protection can potentially last longer than 20 years. Industrial secrets meeting the relevant requirements are protected as long as they remain secret, which is in the hands of the company or the innovator.

A patented invention becomes accessible 20 years after the protection period ends. Therefore, the trade secret option is definitely a better choice for robotic inventions. When this reality exists, it is possible to clearly state that the robotics industry, the greatest innovative power of the 21st century, does not benefit from intellectual property rights protection. This shows once again that intellectual property rights must change within the scope of the new technology age.

As mentioned above, patent protection for robots is not always beneficial. This issue is not actually very new for this field. In the 1980s, most robotic entrepreneurs had many patents that expired before they could profit from the protected products. Another disadvantage of this field is that it takes a long time for robotic innovation to be in use and have market power. In this context, trade secrets may be more beneficial for innovators in this field.

Another issue is that patents still do not provide protection for software, especially computer programs, as they do not meet the requirements. However, if they can be kept as trade secrets, as Google has done for years with its search engine, the importance of trade secrets due to the ineffectiveness of patents in this field can be better understood.

It is known that the integration of software and hardware is the essence of robotics. Robots capable of creating and inventing on their own may soon become common and widely used, like the right to vote resulting from citizenship. If a robot can obtain citizenship from a country, it may create some work as a sculptor. In this case, the question is whether the cost of the sculpture will be paid to Sophia or her owner. After all, there are already robots equipped with AI and can be seen as robot Sophia.

What's in Disruptive Technologies?

Relevant Examples of Artificial Intelligence, Robotics, and Theoretical Discussions

At this point, it is quite important to note that a robotic city is being built. The Neom project has started in Saudi Arabia. In this project, a new city-state will be built with lands from different countries. It will have robotic butlers. Compatible with artificial intelligence, the goods provided by this project will be owned by robots and AIs. Another famous example is Paul, a robot that uses sensors and an arm to draw portraits of people. Inventions by such machines or robots will be seen in the near future. Another example, IBM's Watson, a question-answering computer system, can answer questions posed in natural language. As seen, the difference between humans and robots will decrease over time, and at some point, the property rights of robots will begin to be discussed.

For opposing views, it can always be argued that machines and computers have broader preliminary knowledge than humans in the way they are programmed. However, they are still products of humanity created with human innovation. However, our counter-argument here is that every human being is in contact with another human to come to life. In this argument, robots are also in contact with the humans who created them. The inventions we create, art, music, are not the property of our parents. In this context, it may be a bit challenging, but robots and their creations should also be accepted as their property, not just within the scope of intellectual property rights but also with other legal tools.

The Issue of Ownership of Patent Rights on Inventions Developed by Machines or Robots

The issue of ownership of patent rights on inventions developed by machines or robots must be addressed immediately by the World Intellectual Property Organization and other responsible organizations. This point will probably be vital because innovation is the fundamental issue of our civilization. Whether applications made by computers, AIs, or robots will be accepted as inventors is still a mystery, but as seen, the time for discussion has come and is no longer just a science fiction novel. As mentioned above, the EPO has not yet accepted an AI as an inventor.

Many examples can be added to explain the creativity of AIs. IBM's AI system DeepBlue defeated Garry Kasparov in chess in 1997. IBM's Watson won a Jeopardy game, and Google's DeepMind defeated a Go world champion in 2016. Additionally, DeepMind is better than most human players in Starcraft II, a very popular strategy game. As in the case of Watson, it is also applied to solve problems that humans cannot solve. This has been a productive area where traditionally patentable inventions have been produced.

The Term Computer-Generated Works

On the other hand, computers create works in some cases where no human author, inventor, or composer can be identified. These are called Computer-Generated Works (CGW) or "works created by a computer." Computers have been creating CGWs for decades, but this has not been widely accepted. A very famous example is when technologist Raymond Kurzweil, author of The Singularity is Near, obtained patent protection for a computer program called Cybernetic Poet, which can produce creative writing in 2003.

An AI may consider that its CGW is eligible for intellectual property rights protection. Along with the examples mentioned above, it may be practically challenging to examine whether the works created by the computer were created by the computer itself or as an intermediate work. From the perspective of the examples, challenges, and more mentioned above, it is seen that the foundations of intellectual property rights will change at least due to the definition of inventors. However, it should not be forgotten that other ethical issues must first be reassessed to create a process involving non-human ownership of intellectual property rights, and new property rules within the scope of humanity worldwide must be established.

Europe's View on CGWs

The European Economic Area (EEA) aims to ensure the free movement of goods, capital, services, and labor within the European Union. On the other hand, intellectual property rights such as copyright and patents may pose some obstacles to free trade policy. Intellectual property rights are initially not global and cannot naturally cross borders or be recognized in reverse. The EU has tried to centralize and regulate national intellectual property laws to promote trade. The Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) has been examined by the Court of Justice of the European Union (CJEU) with case law discussed in various EU directives. After TRIPS, several EU directives, such as the Computer Programs Directive and the Database Directive, have been adapted to national intellectual property laws because the rights provided by intellectual property rights include differences in terms of material or duration.

The EPC requires the disclosure of the identities of inventors in patent applications and granted patents but leaves the authority to decide who the inventor is to the laws of the contracting states. From our perspective, this EPC application is also about to change because the definition of identity will need to be reviewed in the near future.

Who Will Own Some Types of Creations in the Future?

In the future, it may even be possible for a computer to identify its own work and decide whether it is patentable. If CGWs are to be protected, the question of who will own the invention and ownership must be immediately asked. This is a moral question rather than an economic one.

This discussion was addressed very extensively and critically by Ryan Abbott in 2017. If CGWs are to be protected and a computer is to be identified as an inventor, it starts with the question of who will own the CGW. According to him, computers should not own patents, but this is still controversial. According to him, computers are not emotional and, therefore, cannot have any ownership, but they are still owned as property by humans. Additionally, Colin Davies's suggestion in Abbott's discussion stated that the computer should own intellectual property rights and transfer them by contract. Colin Davies says this would require machine liability and therefore a deposit on behalf of a computer to satisfy adverse decisions or for an insurance arrangement. This discussion and suggestions continue. The importance of this discussion is that AIs, robotics, or other new technologies encountered bring their challenges, and these issues need to be evaluated and addressed by authorities. Because the coming years will continue to bring new challenges in terms of inventions, technologies, or other things. Machine liability, for example, may be the next problem for intellectual property rights authorities. If the identity issue of CGWs can be resolved, other questions will come as a domino effect.

Following the discussion, it is seen that other options present the possibility that ownership can be directly given to the human computer user. In most cases, these may be the same entity, but they can also be different parties. According to Abbott, the ideal solution may be to give ownership to the party most affected by economic outcomes. According to him, the owner of the computer should be the default owner of any CGW it produces. However, the authorities will face the problem of defining the default owner because many devices are in our hands but belong to a manufacturing company. From one point, this discussion and suggestion may seem logical. On the other hand, it will be no different from owning a child. Because the proposed outcomes are the product of a kind of mental creation, but of a non-human mind. On the one hand, it can also be argued that whoever benefits from this should also take responsibility for it.

The Future: The Rise of Machines

In this section, the need for patents and outdated topics will be evaluated with examples. First of all, it is possible to state that patents could be allocated during the earlier machine invention period. The example of this section is Eurisko, an AI experiment evaluated in a publication in 1982. In this publication, the program is described in the abstract section as follows:

"We apply Eurisko to invent new types of three-dimensional microelectronic devices, which can then be fabricated using newly developed laser recrystallization techniques. Three experiments were conducted, and some new designs and design rules emerged."

Eurisko's work is described as autonomously discovering new knowledge, an expert AI system. It is programmed to work according to a set of known rules but can discover new rules and use them to modify its programming.

To create new microchips, Eurisko was programmed with knowledge and basic rules and evaluation criteria related to basic microchips. This experiment can be considered the first autonomous thinker created by humans and not human.

Various sources suggest that a patent was granted for one of Eurisko's chip designs in the mid-1980s. According to Abbott, sources seem to refer to a patent application for chip design made by Stanford University in 1980. Later, in 1984, the university abandoned it, and no patent was obtained by anyone. At that time, the patent is a very profitable and suitable option because there are many hardware issues that meet the patent requirements in different ways.

Today, machines, robots, especially AIs, are on the verge of a breakthrough. Abbott defines Eurisko's time as the first stage in "Everything is Obvious." According to him, there is a transition from human inventors to machine inventors. This timeline consists of five stages and is very illustrative to understand the past developments of non-human inventors and their possible futures according to current research. The first stage is in the past, where inventors were humans, and the skill standard was also human. The second stage is today, where inventors are still human, but certain AIs (SAI) are also inventors. The advanced person is the person with the needed skills. According to this table, in the short term, in the third stage, humans and SAIs will be seen as equal inventors, and the skill standard will increase as the equality of advanced persons and SAIs. In the fourth stage, inventors will be SAIs and Artificial General Intelligences (AGI) instead of humans. The skill standard of the timeline is advanced AGIs. In the long term, in the fifth stage, it is expected that inventors will only be Artificial Super Intelligences (ASI) and the skill standard will only be ASIs.

These assumptions are quite critical. Because according to this, in the long term, what we know as an inventor will only be human, and inventor will only be an adjective for non-humans. This assumption explains itself explicitly. Finally, according to these assumptions, in approximately twenty-five years, expert opinion will say that the introduction of AGI will start the fourth stage where AGIs will be inventors instead of humans.

Real Life to Possible Future

Companies like Google, IBM, Apple, Amazon, Facebook, even Uber, have some types of invention machines ready and waiting within the scope of intellectual property. Most are kept as trade secrets but are known, and their capabilities are featured in the news almost every day. These companies have versatile technology, from wearable devices to super complex algorithms. If the invention step of patent law regarding intellectual property is disrupted, these big businesses will probably be the actors of it. Because although such challenging technologies require less investment as days go by, significant investments must continue to push authorities for change.

Update in 2024: Large Language Models and ChatGPT

Large Language Models and Intellectual Property Rights

As of 2024, large language models (LLMs) and AI applications like ChatGPT significantly impact intellectual property rights. Large language models are AI systems trained on massive amounts of data and capable of performing complex language tasks. These types of AIs can generate text, translate, summarize, and perform many other language-based tasks.

Large language models like ChatGPT enable the quick and efficient creation of text-based content. This provides a great advantage for writers, researchers, journalists, and other content creators. However, the impact of these technologies on intellectual property rights cannot be ignored.

Issues and Solutions Regarding Intellectual Property Rights

The ownership of content generated by large language models and how to protect this content have created new problems in the field of intellectual property rights. For example, who owns a text generated by ChatGPT? Should this text be considered the work of a writer, or should it be categorized differently because it was created by AI?

Such questions require the reevaluation of intellectual property law. New legal regulations and definitions are needed to protect the works created by AI. Specifically, clear rules must be established on whether the works produced by AI will be protected under copyright and how this protection will be applied.

The Use of ChatGPT and Other AIs

The use of ChatGPT and similar AIs is not limited to content creation. These technologies are used in a wide range of areas, from customer service to educational materials, from healthcare services to financial analysis. This emphasizes the impact of AI on society and the necessity of managing these impacts.

The rapid development of AI technologies requires the legal framework to keep up with this speed. Large language models like ChatGPT have the potential to revolutionize many fields with their ability to generate human-like texts. However, it should be remembered that this potential can also create ethical and legal issues.

Assessment of the AI Act Coming into Force in the European Union

Key Elements of the AI Act

The AI Act stipulates the rules that must be followed during the development, distribution, and use of AI systems. This bill classifies AI systems according to their risk levels and brings different regulations for each class. Stricter regulations are foreseen for high-risk AI systems, while more flexible rules are applied to low-risk systems.

The main objectives of the AI Act include:

• Safety and Transparency: Ensuring the safe and transparent use of AI systems. This will enable users to understand how AI systems work and what kind of data they use.

• Ethics and Fairness: Promoting the ethical and fair use of AI systems. This includes preventing discrimination and respecting human rights.

• Responsibility and Accountability: Establishing mechanisms of responsibility and accountability for the harm caused by AI systems. This will encourage the safe and responsible development of AI systems.

The AI Act's Impact on Intellectual Property Rights

The enforcement of the AI Act will also lead to significant changes in the field of intellectual property rights. New regulations will be required regarding the ownership of content created by AI and the protection of this content. The AI Act aims to ensure the protection of intellectual property rights during the development and use of AI systems.

• Patents and Copyrights: Patent and copyright protection will be provided for inventions and works created by AI systems. However, clear rules must be established on how this protection will be applied and to whom the content created by AI will belong.

• Trade Secrets: The AI Act will also bring important regulations on the protection of trade secrets. Trade secrets developed by AI systems can be more effectively protected with this law.

• Ethical and Legal Responsibility: The ethical and legal responsibilities of AI systems will also be evaluated within the scope of intellectual property rights. The ethical and legal responsibilities of content created by AI will directly affect the protection and use of this content.

Conclusion

Even today, the solutions to the above-mentioned problems do not exist and await resolution by the authorities. To accelerate this, it is seen that the number and quality of invention machines need to increase. The impact of the 21st century and disruptive technologies will also be seen in our definitions of industry. The support behind the idea of financial gain can turn into knowledge accumulation or other non-economic terms as we know them today. The term economy itself may change and should be questioned within the scope of intellectual property. As we know today, the EPO does not want to evaluate invention machines or AIs as inventors, but in the near future, they will need to present their arguments based on legal foundations. The only obstacle in front of these changes is the definitions we have created.

The AI Act brings important regulations to ensure the safe, ethical, and responsible use of AI technologies. This law will ensure the protection of content created by AI under intellectual property rights while also promoting the safe and fair development of AI systems. With the AI Act coming into force, the impact of AI technologies on society can be better managed, and the innovations brought by these technologies can be more effectively evaluated within the framework of ethical and legal responsibilities.

In short, while 3D printing technology changes our definitions of industry and economy, the purpose of production also transforms from the ability to create something from nothing to becoming human with invention machines and AIs. Today, humanity is on the verge of giving the unique feature called inventor to machines. Some breakthroughs like AI, three-dimensional printers, quantum computing systems, and these CGWs/invention machines, genetically modified organisms have already entered our lives, and some are on the way, such as androids, cities exclusively for robots, AI inventors, and more. For the reasons stated in this assessment, the reorganization of intellectual property terms is essential and inevitable.

References

• Websites

  • Robotticelli: the mechanical marvel creating extraordinary works of art

  • Saudi Arabia citizen robot

  • Elon Musk Neuralink brain chip

  • IBM Watson for Drug Discovery IBM

  • [DeepMind's StarCraft 2 AI better than 99.8 percent of all human players](https://www.theverge.com/2019/10/30/209 39147/deepmind-google-alphastar-starcraft-2-research-grandmaster-level)

  • RoBotticelli: the mechanical marvel creating extraordinary works of art (8 September 2015)

  • It's being built on our blood: the true cost of Saudi Arabia's $5bn mega-city Neom (4 May 2020, Ruth Michaelson, The Guardian)

• Online Journals and Articles

  • European Patent Office Declaration 28 January 2020

  • Convention on the Grant of European Patents of October 1973

  • Understanding Industrial Property World Intellectual Property Organization (Retrieved 2018-12-06)

  • Heuristic Search for New Microcircuit Structures: An Application of Artificial Intelligence (1982, Douglas B. Lenat, William R. Sutherland, James Gibbons, AI Magazine)

  • Everything is Obvious (2019, Ryan Abbott, UCLA Law Review, 66(1))

  • Economic implications of 3D printing: Market structure models in light of additive manufacturing revisited (2015, Christian Weller, Robin Kleer, Frank T. Piller, International Journal of Production Economics, Volume 164)

  • An evolutionary step in intellectual property rights- Artificial intelligence and intellectual property (2011, Colin Davies, Computer Law and Security Review, 27:601-615)

• Legal Sources

  • Directive (EU) 2016/943 of the European Parliament and of the Council of 8 June 2016 on the protection of undisclosed know-how and business information (trade secrets) against their unlawful acquisition, use, and disclosure (Text with EEA relevance, OJ L 157, 15.6.2016)

  • Paris Convention for the Protection of Industrial Property of March 20, 1883; Patent Cooperation Treaty done at Washington on June 19, 1970

  • Du Pont/Appetite Suppressant (ECLI:EP:BA:1986)

  • The guidelines for Examination

• Other Sources

  • Patents or Trade Secrets: The Choice Is Yours (Michael R. McGurk, Rachel L. Emsley)

  • Artificial Intelligence, Big Data, And Intellectual Property: Protecting Computer-Generated Works In The United Kingdom (Forthcoming 2017, Ryan Abbott, Research Handbook on Intellectual Property and Digital Technologies, Tanya Aplin ed, Edward Elgar Publishing Ltd)