Neural Network Copyright: 5 Key Points to Understand

Contents:

• Copyright: Essence and Ownership
• Definition of Artificial Intelligence in Legislation
• Authorship of Objects Created by Neural Networks: Who Holds the Rights?
• The Practical Side of the Issue: What's the Situation with Patents and Copyrights for AI?
• Current Conclusions on the Legal Status of AI

Copyright: Essence and Ownership

Copyright is a complex and multifaceted concept that is not always clearly defined. These rights protect creative works, including literature, music, visual art, and other forms of expression. It is important to understand that copyrights can vary depending on jurisdiction and specific circumstances. This makes the issue of intellectual property protection relevant and requires a careful approach to its study.

Copyright and authorship are two fundamental legal concepts in the field of intellectual property. These categories determine who has the right to protect their works and how the created content can be used. Copyright provides owners with the ability to control the use of their works and benefit from it, while authorship secures authorship and protects the personal interests of the creator. Understanding these concepts is important for both individual creators and organizations working with intellectual property.

Property copyrights typically belong to the creators of works. In different jurisdictions, these rights may belong to individuals, such as authors and co-authors, or legal entities, such as organizations. This is especially important for works created for hire, such as software or media articles. Properly formalizing property copyrights protects the interests of creators and ensures the lawful use of their works.

In Russia, copyright belongs to the employer by default. To retain rights to your works, you must enter into a special agreement. The situation becomes more complicated if you work under a civil law contract (CLC) or are self-employed. In such cases, the property rights remain yours, but the terms of the contract with the client may automatically transfer these rights to them. Therefore, it is important to carefully review the terms of all agreements and make changes if necessary to protect your copyright.

Each country has unique legal features regarding copyright. In the United States and the United Kingdom, ownership of a work often rests with the organizations that have made a significant contribution to its creation. Under UK law, copyright is viewed as a property right, which emphasizes the importance of legally protecting works and the interests of their creators. This difference in approaches to copyright affects the distribution of rights and responsibilities between authors and organizations, which is important to consider when creating and using creative works.

Authorship, as opposed to property rights, usually belongs to individuals. In the United States, the law clearly states that an inventor can only be an individual or a group of people. This emphasizes the importance of human contribution to the process of creation and innovation. Attribution protects creative achievements and ensures that creators are recognized.

European Union Directives 2009/24 and 96/9 establish authorship as an exclusive right of individuals. This position is also supported by the Russian Civil Code, which defines the author of a work as the person who directly created it. Establishing copyright plays a key role in protecting intellectual property and promotes creative activity.

In China, it is possible to recognize authorship for both individuals and legal entities. This is similar to the Soviet practice, under which the author of works such as photographs or films could be not only an individual but also an organization. This system allows for the contribution of both individual creators and entire groups to be taken into account, ensuring copyright protection in various fields of art and culture.

In the United States, an interesting precedent arose when the law did not specify that the author of a work must be exclusively human. This case led to a lawsuit concerning photographs taken by a monkey. British photographer David Slater, who captured the macaques, found himself at the center of a dispute with PETA, which claimed that the macaque itself was the author of the photographs. This case raises important questions about intellectual property ownership and the status of animals under copyright law.

After eight years of litigation, the U.S. Copyright Office ruled that only a human can be considered the author of an original work. This decision underscores the importance of the human element in the creative process and its significance for copyright protection.

Lawmakers have definitively clarified that the U.S. Copyright Office will only register original works if they are created by a human.

Read more:

The history of copyright covers a long path, from ancient times to modern digital technologies. In ancient times, such as the era of the creation of the Iliad, works of art and literature did not have formal legal status. However, even then, ideas about the protection of intellectual property existed. Over time, with the development of society and culture, the need arose to establish laws regulating the rights of authors.

In the Middle Ages, the advent of the printing press led to a new stage in the history of copyright, when books began to be distributed en masse. This contributed to the formation of the first copyright laws, such as the Royal Privilege Act in England. Since the 18th century, with the adoption of the Copyright Act, authors began to receive formal rights to their works.

In the 20th century, with the growth of technology and the emergence of new forms of media such as radio, television, and the internet, copyright issues became even more pressing. In response, international agreements such as the Berne Convention were adopted, providing copyright protection on a global level.

In recent decades, the development of open-source technologies such as GNU and Linux has raised new questions about the nature of copyright. These projects promote the ideas of free access and collaborative creation, which requires a rethinking of traditional approaches to intellectual property protection.

Thus, the history of copyright is a dynamic process, reflecting changes in cultural and technological contexts, which continues to evolve in the digital age.

Definition of Artificial Intelligence in Legislation

Artificial intelligence (AI) began to be officially recognized in Russian legislation in 2019, when it was first mentioned in a Presidential Decree. In 2020, this definition was enshrined in Federal Law No. 123-FZ, marking an important step in developing the legal framework for AI technologies. Legislative recognition of AI opens new prospects for its implementation and use in various fields, such as economics, healthcare, and education. Legal regulation of AI contributes to the creation of a safe and ethical environment for its use, which is relevant in the context of rapid technological progress. According to Article 2, Clause 1 of Federal Law No. 123-FZ, artificial intelligence is defined as a set of technologies that imitate human cognitive functions. This includes the ability to self-learn and make decisions without predetermined algorithms. The results of artificial intelligence are comparable to those of human intellectual activity. Understanding this technology is important for assessing its impact on various areas of life and business. Artificial intelligence is used in medicine, finance, education, and many other industries, highlighting its importance in the modern world.

In 2021, the United States passed the National Artificial Intelligence Initiative Act, which clearly defines the concept of artificial intelligence. According to this law, artificial intelligence is a machine system capable of achieving human-defined goals, making predictions, providing recommendations, and making decisions that affect the physical or virtual environment. This definition underscores the importance of AI in modern technology and its potential impact on various areas of life.

In March 2023, the European Parliament approved a draft almost identical to the American definition. According to the draft EU Act on Artificial Intelligence (AI), "an AI system is a machine system designed to operate at varying levels of autonomy, capable of generating outputs such as predictions, recommendations, or decisions that affect the physical or virtual environment." This document highlights the importance of regulating AI technologies in Europe, ensuring a balance between innovation and security.

AI and the neural networks that power it are technologies protected by intellectual property rights. These include copyright, patents, and trade secrets. These legal instruments play a key role in protecting innovation and creativity in AI, thereby facilitating the development and adoption of new technologies.

AI-based software is subject to the same principles as any other software. It can be either proprietary or open source. Proprietary solutions provide users with access to functionality but restrict the ability to modify and distribute the code. Open source software, on the other hand, allows developers and users to freely modify and distribute the code, fostering collaboration and innovation in AI. The choice between these types of software depends on user needs and development goals.

Authorship of Objects Generated by Neural Networks: Who Owns the Rights?

The topic of interpreting artificial intelligence (AI) as a useful tool or something more is a topic of active debate. A key aspect is the human influence on the final result created by AI. It is important to consider that AI is capable of processing and analyzing large amounts of data, which significantly facilitates work in various fields, such as medicine, finance, and marketing. However, despite its capabilities, the final decision and responsibility remain with humans. Thus, understanding the role of AI in modern society requires a thorough analysis of its potential and limitations, as well as an awareness of the need for human oversight.

A neural network that generates code is a powerful tool in the field of technology. However, when it independently creates code, the problem of determining copyright arises. This process lacks human creativity, which calls into question the protection of such works. It is important to consider that the rights to the created code can be contested, since authorship in the traditional sense does not exist. This raises questions about legal regulation and the need to adapt legislation to the new realities associated with the use of artificial intelligence in software development.

The main difference between artificial intelligence and a person working under an employment contract is that AI acts as an object of law, not a subject. This situation raises an important question for lawyers: should the status of AI in the legal field be reconsidered? The development of technology and the increasing use of artificial intelligence in various fields of activity pose new challenges for the legal system, requiring careful analysis and possible changes to legislation.

The question of whether the results obtained by artificial intelligence without human participation can be recognized as an object of intellectual property is relevant and requires careful analysis. If the answer is yes, it is necessary to determine who will own the corresponding rights to these results. Legal experts are exploring several possible scenarios, including whether rights belong to the AI ​​developers, the users who initiated the process, or the AI ​​system itself. This issue requires further research and discussion to ensure appropriate legal protection and a fair distribution of rights to the results of creative work generated using new technologies.

• the programmer who developed the software;
• the user of the AI ​​program;
• the owner of the equipment used to run the program;
• the AI ​​itself.

There are various scenarios in which the results of artificial intelligence could enter the public domain. This could be beneficial for both creators and ordinary users of neural networks, since the generated objects will be available for free use. However, for large companies, this prospect raises concerns. AI objects are becoming increasingly sophisticated and have high commercial potential, which makes their protection and exclusive use important for business. In an increasingly competitive environment, companies are seeking to maintain control over their developments in order to extract maximum benefit from innovative technologies.

Modern technologies make it possible to replace designers, programmers, and journalists with software. However, legal protection is necessary to protect products created using artificial intelligence. This highlights the importance of developing new legal norms and approaches that take into account the specifics of AI-assisted content creation and use. Legal protection for such products will ensure the interests of not only creators but also users, facilitating the development of innovative solutions in various fields.

Resolution 2020/2015 of the European Parliament on intellectual property rights in the context of artificial intelligence technologies distinguishes two types of works: those created by humans using AI, and fully autonomous works created by artificial intelligence. Currently, the latter do not have legal protection under copyright law. This raises important questions about the status and protection of works created without human intervention and requires further discussion and the development of legislative initiatives in the field of intellectual property.

The resolution emphasizes that granting artificial intelligence legal personality is inappropriate. This could have a negative impact on the development of human creativity and innovation. It is important to consider that AI, although capable of performing complex tasks, cannot replace the unique abilities and creativity of humans. Maintaining the primacy of human creativity is key to ensuring sustainable progress in society.

New technologies have the potential to infringe the rights of authors of original works used to train and create new artificial intelligence objects. In 2020, a report was prepared for the European Commission entitled “Trends and Developments in AI. Challenges for Intellectual Property Law”. This document identifies four criteria that determine the protection of works created using AI under EU copyright law. These criteria underline the importance of protecting intellectual property in the context of rapid technological developments and their impact on the creative process.

• The work belongs to the literary, scientific or artistic field;
• A human being has applied his or her intellectual effort;
• The work demonstrates the author’s creative choice;
• This choice is expressed in the final result.

The report identifies three main stages of the creative process. The first stage is idea generation, when free thinking and concept development take place. The second stage involves selecting and refining ideas, which allows us to identify the most promising ones. The third stage is implementation, during which ideas are transformed into specific products or projects. Understanding these stages helps optimize the creative process and improve work results.

• Concept;
• Execution;
• Editing.

When artificial intelligence is involved in the content creation process, the primary human role is retained in the first two stages. The author forms the general concept and sets the design of the work, which allows them to direct the further process. The person makes a creative and free choice, which is later realized in the final result. It is also important to note that it is at this stage that editing occurs, which helps bring the work to the desired quality. Thus, artificial intelligence acts as a tool that complements and supports human creative efforts, but does not replace them.

To determine the authorship of works created using neural networks, it is necessary to carefully evaluate the human contribution and their creative choices. It is important to consider that authorship is not limited to the technical aspect of creation. The degree of human involvement in the process, their ideas, concept, and individual style should be analyzed. This will allow us to more accurately determine where the influence of technology ends and the author's unique creativity begins. Establishing clear criteria for assessing authorship will help protect the rights of creators and ensure fair use of works created using neural networks.

• The degree of human intervention;
• The level of autonomy of AI systems;
• The origin of the data used by AI and its legal protection.

Let's consider the hypothetical case of a programmer using the ChatGPT neural network to write code. According to current legislation, only code created using the programmer's personal creative work can be considered an author. At the same time, the code generated by ChatGPT is not subject to copyright protection. This raises important questions about the legal status of works created using artificial intelligence and how the contribution of neural networks to the creative process should be taken into account. It's important to understand that the use of neural networks in software development challenges traditional notions of authorship and intellectual property. If a developer fails to disclose that part of their code was created using artificial intelligence, this can raise legal and ethical issues. Because the rights to use the technology and its results remain unclear, users may be considered the authors of the code, as ChatGPT and similar tools are designed to meet their needs. It's important to consider that transparency in the use of AI in software development can reduce the risk of legal disputes and increase trust among users. Developers are encouraged to clearly indicate when they use AI in their projects to avoid misunderstandings and ensure copyright compliance.

If a program includes code fragments protected by different licenses, there is a potential for claims from their authors or copyright holders.

The practical side of the issue: what is the situation with patents and copyrights for AI?

The state of the intellectual property legal system has raised a significant number of questions in recent years, especially in light of the rapid development of artificial intelligence technologies. A striking example of this is Dr. Steven Thaler's attempt to obtain patent rights for the DABUS AI, which independently developed two innovative products: an improved food container and an emergency warning light. This case raises important questions about the legal status of AI as a creator and how modern intellectual property laws can adapt to the new realities associated with developing technologies. It is important to consider how legislation can be revised to take into account advances in AI and ensure the protection of the rights of both inventors and the technologies themselves.

Thaler, inspired by the success of the DABUS system, decided to file patent applications in which he indicated artificial intelligence as the author. However, most patent offices, including those in Canada, Germany, and the United States, expressed skepticism about the initiative and rejected all but one application, which was approved in South Africa. This event highlights the relevance and complexity of issues related to intellectual property rights in the era of rapid development of artificial intelligence technologies.

Thaler continued his efforts, and in July 2021, an Australian court made a historic decision recognizing artificial intelligence as legally capable of being an inventor. This decision became a first in global practice, opening new horizons for the legal regulation of intellectual property in the context of technology. The significance of this event lies in its challenge to traditional notions about the role of humans in the invention process and may change the approach to patenting innovations created with the help of AI.

The joy from the decision was short-lived: on April 14, 2022, the Federal Court of Australia overturned this ruling, citing the Patents Act 1990. According to this law, only a legally competent natural person can be an inventor. The court noted that inventions are formed in the human mind and require their active participation.

In 2022, the case of "Artists v. Midjourney" arose, when a group of authors filed a lawsuit against the company. The artists claimed that the neural network was merely combining and modifying their works, without holding the rights to use these materials. This case raises important questions about creative ownership in the age of artificial intelligence and how technology impacts copyright.

The US Copyright Office responded to the lawsuit, siding with the authors and canceling the registration of images created with Midjourney. This decision highlights growing concerns about copyright protection in the era of artificial intelligence. With the active use of AI technologies, the importance of copyright enforcement is becoming increasingly pressing, necessitating a review of legislation and practice in this area.

There are also positive examples in the field of copyright for content created with artificial intelligence. In 2020, a court in Shenzhen, China, ruled in favor of the owner of an article created by the chatbot Dreamrighter. The court noted that the text was original and structured, which complied with copyright requirements. This case highlights the importance of recognizing and protecting intellectual property, even when content is generated using artificial intelligence technologies.

The situation with artificial intelligence rights and copyright is complex and multifaceted. Legislative uncertainty creates both new opportunities and significant challenges for creators and developers. It is important to closely monitor changes in the legislative framework and case law in order to adapt to new conditions and avoid legal risks. Advancement in the field of AI requires understanding existing regulations and actively participating in the discussion of their changes, which will help shape clearer rules of the game for all participants in this rapidly evolving market.

Current Findings on the Legal Status of AI

The legal status of works co-authored with artificial intelligence is becoming a hot topic for legal and technology specialists. This issue requires in-depth analysis and discussion, as legal regulations do not always keep pace with rapidly evolving technologies. It is important to explore how copyright may apply to creations involving AI and what legal mechanisms can protect the interests of all parties.

With the rapid development of neural network technologies, significant changes in copyright law are expected in the coming years, both in Russia and internationally. Key aspects requiring clarification include questions of ownership of generated content, liability for copyright infringement, and the need to adapt existing regulations to new realities. It is important to consider that neural networks are capable of creating original works, which challenges traditional approaches to copyright. The need to update legislation is becoming increasingly urgent to protect the rights of content creators and ensure fair remuneration for the use of intellectual property.

• Should AI-generated content be protected by intellectual property?
• If so, which legal regime is best?
• Can a person not actively involved in the creative process be recognized as the author of such products?
• Should AI be considered a legal entity and granted authorship rights?

It is important to consider how these changes may impact labor legislation. This issue requires careful research and further discussion to ensure compliance with new requirements and the protection of workers' rights.

For a more complete understanding of this topic, we strongly recommend that you read the additional materials. This will deepen your knowledge and expand your understanding of the issue under consideration. Reading specialized sources will help you better understand the key aspects and current trends in this field.

• What is a neural network and how does it work?
• Test: can neural networks achieve the same results?
• OpenAI: the rise of the company that developed ChatGPT