This article has been written by Anjali Jain pursuing a Diploma in Law Firm Practice: Research, Drafting, Briefing and Client Management course from LawSikho.
This article has been edited and published by Shashwat Kaushik.
Table of Contents
Introduction
The rapid advancement of generative artificial intelligence (AI), providing exciting opportunities for innovation in all sectors of the economy, brings serious questions related to copyright infringement, fair use of existing works and ownership rights. Vast datasets, which may include copyrighted material along with licensed ones, are used to train generative AI systems such as OpenAI’s GPT-4 to generate new original content such as images, music, codes, texts, and other forms of content that closely resemble human-generated works. Copyright law protects creators’ rights by granting them exclusive control over their works reproduction and distribution. However, the rise of generative AI challenges traditional notions of originality and authorship.
The main focus of this article is to analyse copyright violations by generative AI. It will deal with key issues such as ownership of AI-generated content, use of copyrighted material in training data, fair use doctrine, relevant case laws and precedents and international perspective. By examining the above aspects related to generative AI, we can find a better way to guide and develop generative AI in such a way that no intellectual property rights are damaged by its use.
The nature of generative AI
Generative AI is different from traditional AI as it doesn’t classify old data but creates new and original content such as texts, images, music, codes, and more after being trained by large datasets. For instance, if you write a proper prompt to “Write a story about a lost person,” GPT-4 can create a unique story, including dialogue and plot, based on learnt patterns. This ability to produce creative and innovative content is great but sometimes generated content resembles existing works, which may raise ethical and legal concerns.
Fundamentals of copyright law
Copyright law protects the rights of original creators by granting them several exclusive rights over their creations to enable them to earn profit from those, to control how their work is used and to motivate people to do more innovative and creative work. These rights include reproduction, distribution, and creation of derivative works based on original work. Copyright law applies to intellectual creations such as literary, musical, or artistic creations.
Copyright is an important intellectual property right in accordance with the report of TRIPS. In India, copyright is governed by the Copyright Act, 1957, which was amended in 2012. Section 14 of this Act grants creators of original works exclusive rights to publish and broadcast their creative works.
To get copyright protection, authors’ work should be original with some degree of creativity and it must be fixed in a tangible medium, such as written on paper or digitally stored. Though no formal registration is required, if registered, it gives certain legal benefits.
Copyright infringement
Copyright infringement in India is primarily governed by Sections 51 and 52 of the Copyright Act, 1957, which outline certain actions constituting infringement and certain exceptions, such as fair use. When someone uses copyrighted work of some owner without his permission, then it violates the rights of the owner of such copyright unless it’s covered under exceptions as stated under Section 52 of the Copyright Act. For instance, downloading and sharing copyrighted movies without the owner’s permission is infringement, as it violates the owner’s exclusive rights to distribute and earn profit from their work.
Generative AI and originality
Defining originality
Originality is a fundamental principle in copyright law that serves to protect the unique and creative expressions of authors. It involves the concept that a work must be self-created and bear the unique imprint of the author’s personal ideas, thoughts, and imagination. Originality does not require a work to be groundbreaking or revolutionary, but it should demonstrate a certain level of creativity and distinctiveness.
It is essential to distinguish originality from mere imitation or copying. Simply copying existing content or making superficial changes to it does not constitute original work. The author’s contribution must be substantial and reflect their unique perspective and style. This includes elements such as the selection, arrangement, and expression of ideas, as well as the author’s personal interpretation and interpretation of the subject matter.
To assess originality, copyright law considers factors such as the nature of the work, the author’s intent, and the degree of skill and effort involved in creating the work. The level of originality required may vary depending on the type of work, with creative works like novels or paintings requiring a higher degree of originality compared to functional works like user manuals or technical drawings.
Originality is crucial in copyright law because it ensures that authors are recognised and rewarded for their unique contributions to society’s cultural and intellectual heritage. It encourages creativity and innovation by providing authors with the incentive to create new and original works. Without the protection of originality, authors would be more likely to resort to copying existing works, stifling creativity and limiting the diversity of creative expression.
Eastern Book Company vs. D.B. Modak is a very important case that set the standards for originality in India. In the above case, EBC claimed copyright infringement against software copying elements from its journal, SCC, by claiming that SCC does have any copyright over its content because government documents are not copyrightable, thus judicial decisions shouldn’t be copyrightable. To test “originality,” various standards were proposed but the Supreme Court rejected the “Sweat of the Brow” as it was too low a standard and the “modicum of creativity” as it was too high a standard to decide copyright issues and finally adopted the Canadian test, which requires the work to reflect the author’s skill and judgement. The court gave the final verdict in favour of EBC, claiming that the additional elements in the SCC version of judgements involved sufficient human creativity to be called “original” in accordance with the Canadian test. The court also reinforced that outputs from simple prompts lack the necessary criteria of originality for copyright protection.
Challenges to originality with AI-generated content
Generative AI leads to different challenges to the concept of originality. Outputs from AI models may resemble existing works, which ultimately raises doubt regarding their originality. Additionally, traditional criteria of assessing creativity do not easily apply to machine generated content, creating legal uncertainty regarding copyright protection and ownership issues for AI-generated works.
Authorship and ownership
Traditional notions of authorship
Traditionally, authorship in copyright laws is given to the creator of original works. This notion only provides exclusive control over intellectual property to human creators. In India, it is emphasised by giving exclusive control to the creator of the work.
AI as an author: legal and practical issues
The rise of generative AI raises questions on conventional authorship and ownership concepts. According to Section 17 of the Indian Copyright Act, only natural persons are given ownership rights. As AI-generated content is autonomous, questions arise about their status as authors. It’s still not clear if AI-generated work should be allowed copyrights or not. And if allowed, who should be given authorship rights: program, programmer, or user?
Programmer vs user : who owns the output
Determining ownership of AI-generated output is a complicated process. Both the programmer who developed the AI and the user who creates a unique prompt that ultimately generates content may assert their ownership rights. The ‘Terms of Use’ of the specific AI may help in deciding this question. The Indian Copyright Act doesn’t deal with this dilemma, leading to ongoing legal debates regarding the rightful owner of AI-generated content. “Sweat of the brow” doctrine is used in many places to decide this question on the basis of effort and diligence put in by the creator.
“Zarya of the Dawn,” a pivotal case involving artificial intelligence (AI) and ownership rights, has emerged as a trailblazer in the ever-evolving landscape of copyright protection. The core of this case revolves around a captivating comic book adorned with AI-generated images, sparking a debate over who rightfully owns the copyright to these digital creations.
Initially, the U.S. Copyright Office, a federal agency responsible for the registration and protection of copyrights, denied copyright protection for the AI-created images, citing the absence of human authorship as the primary reason. This decision sent shockwaves through the creative community, raising concerns about the legal status of AI-generated works and their potential impact on traditional notions of copyright ownership.
However, the creators of “Zarya of the Dawn” refused to accept this setback. They appealed the Copyright Office’s decision, arguing that the AI system used to create the images was merely a tool, much like a paintbrush or a camera, and that the artistic vision and creativity behind the comic book originated from human minds.
This case has the potential to set a precedent for future copyright claims involving AI-created works. A ruling in favor of the creators of “Zarya of the Dawn” could establish that AI-generated images can indeed be protected by copyright, provided that they meet the requisite standards of originality and human authorship. Conversely, a decision upholding the Copyright Office’s denial could create a significant barrier to copyright protection for AI-generated works, making it more challenging for artists and creators to secure legal recognition and protection for their AI-infused creations.
The outcome of “Zarya of the Dawn” is eagerly anticipated by legal experts, technology enthusiasts, and artists alike. Its implications extend beyond the realm of copyright law, touching upon broader questions about the nature of creativity, the role of AI in artistic expression, and the future of intellectual property rights in the digital age.
As the case progresses through the legal system, it will undoubtedly shape the trajectory of copyright protection in the face of rapidly advancing AI technologies. “Zarya of the Dawn” serves as a reminder that the intersection of law and technology often presents complex challenges, requiring careful consideration and thoughtful examination to ensure that the delicate balance between innovation and the protection of intellectual property is maintained.
Infringement by generative AI
Generative AI’s ability to create content using training datasets, which may also include copyrighted material along with licensed and publicly available data, is a huge concern when it comes to copyright protection issues. Deciding factors include the final use of content along with the fact whether AI output infringed on any existing work or not.
The Eastern Book Company vs. D.B. Modak case, as explained above, provides a foundation for understanding the court’s approach towards originality and authorship issues for AI-generated content in India. But it doesn’t provide a proper way that can be used to determine such issues. There is a need to establish proper laws that will deal with all aspects of this issue. With advancement in technology, new legal precedents will come into place to address this issue in a proper way and maybe specific laws might come into the picture to address this challenge in detail.
The US Congress has already introduced a new bill to force AI companies to showcase a detailed list of copyrighted material used for training data as input for the creation of unique output in accordance with the prompts entered by their users. The Indian government should also start addressing this concern with stringent actions.
Till date, many class action suits have been filed against generative AI companies for copyright infringement by famous authors like John Grisham, George R.R. Martin, comedian Sarah Silverman, and many others. With time, cases are only growing and there is no proper law dealing specifically with such cases. Few such cases include The New York Times accusations against OpenAI and Microsoft for using its articles without ownership rights, Universal Music Group’s accusations against Anthropic for illegally using its song lyrics or Stability AI being sued for copying images by Getty. On the other hand, OpenAI, Meta, Microsoft, and a few other companies argue that generative AI benefits society. Their arguments are supported by academic theories and international precedents. They claim that the data used for training is lawful under the fair use doctrine.
Use of training data
The use of copyrighted material for training AI models is a controversial issue. AI systems are trained on vast human-created datasets, such as books, images, and music, without getting a proper license for the same. For instance, GPT-3.5, a large language model, is trained on over 570 GB of filtered text data, equivalent to around 300 billion words or 1.3 million books, exceeding three times the text in the Library of Congress.
Trained models generate unique and creative output, as per the prompt entered by the user, after analysing the input data. Training data often includes unlicensed copyrighted works, raising serious questions about violations of copyright law. To address these issues, several factors need to be studied. These include the application of the fair use doctrine, whether consent of the original creators has been taken or not and a few others.
In the case of Authors Guild vs. Google, Inc., Guild sued Google for copyright infringement after the launch of the Google Books project. Millions of books were scanned to create a searchable library in this project. But the court emphasised that the searchable database by Google will not replace original books because it displays only limited snippets of the books, which enhance data accessibility without harming the actual market of the original books, leading to fair use and not copyright infringement. Thus, it provides benefit to the public at large with its transformative use without violating the original owner’s rights. Hence, the court ruled in favour of Google. This case laid down certain conditions, which, if fulfilled, then training of data is possible without infringing the rights of the original authors. It is still a grey area that needs proper legal consideration.
Fair use doctrine and AI
The fair use doctrine is used to answer the vital question, i.e., whether use of unlicensed data as training data for AI is right or not. It allows limited use of unlicensed copyright material if it lies within four established balancing factors, which are purpose, nature, amount, and market effect. This concept of fair use was first established in the US in the year 1841 during the case of copyright infringement of the writings of George Washington.
Using data to train AI models might not be infringement because it has a transformative purpose to it but the final outcome of that trained data may lead to copyright infringement, which is to be decided by applying the fair use doctrine as given under Section 107 of the US Copyright Act. When a work is used for purposes such as criticism, commentary, news reporting, teaching, or research, it is generally not considered copyright infringement. Fair use is determined by four specific criteria.
- The nature and intent of the use: If the material is used for non-profit purposes such as education or research purposes, it may come under fair use, but if it is used for commercial purposes without any permission, it may be considered infringement. For instance, if you use a copyrighted poem to teach a class, it’s fair use but if you use it as your own creation in a show for money, it is copyright infringement.
- The composition of the protected work used: Creative work can be fictional or factual. Non-fiction or factual work such as legislation is more likely to support a fair use claim than fiction work such as fiction novels.
- The proportionate size and significance of the work used: It is important that the work used is not the heart of the project. For instance, reusing a sentence from a poem is more likely to be fair use than an entire paragraph. Though the entire work is used as training data but is not publicly accessible, thus covering this criterion as claimed by Open AI.
- The measure of the impact on the potential market and value of the original copyrighted work: It’s one of the most important factors to decide fair use. For instance, making copyrighted material publicly available will decrease the value of original work. In the case of generative AI, copyright owners contend that their creations are used in a way to undermine their economic opportunities, thus it is not fair use.
Substantial similarity analysis
Generated AI uses training data from various sources as its input in order to produce output that may resemble the existing original work of authors, which was used as input. If the degree of similarity of AI-generated output to the original copyrighted content is greater, then it may constitute an infringement as it is considered a copy and not original work. This practice of measuring degree of similarity is called Substantial Similarity Analysis. It is used as a measure to decide copyright infringement issues.
The Eastern Book Company vs. D.B. Modak case, a landmark decision in Indian copyright law, brought to the forefront the fundamental requirement of human skill and judgement in determining the originality of a work. The case established a crucial precedent, emphasising that merely making changes to an existing work is not sufficient to qualify it as an original creation. This principle has far-reaching implications for artificial intelligence (AI).
If we apply the precedent set in the Eastern Book Company case to AI-generated works, it becomes evident that simply copying or imitating existing works without substantial transformation would constitute a clear case of copyright infringement. AI-generated works that lack creativity, originality, and human intervention may not be considered original works entitled to copyright protection.
The Eastern Book Company case highlighted the importance of human skill and intellect in the creation of original works. The court recognised that even if certain changes or modifications are made to an existing work, it does not necessarily transform it into an original creation. For a work to be considered original, it must demonstrate a unique and distinctive expression that is attributable to the author’s own creative endeavour.
Applying this principle to AI-generated works, it becomes clear that AI systems, while capable of producing impressive results, lack the inherent human qualities of creativity and originality. AI algorithms are designed to process and manipulate data based on predetermined parameters and instructions. While AI can generate content that may appear novel, it often lacks the personal touch, the subjective interpretation, and the emotional depth that are characteristic of human-created works.
Therefore, in light of the Eastern Book Company precedent, AI-generated works that are substantially similar to existing works without sufficient transformative elements could potentially be deemed as copyright infringements. To avoid such infringements, it is crucial for AI developers and users to prioritise originality and creativity in the development and application of AI systems. This may involve incorporating unique human input, allowing for user customisation, and encouraging AI systems to explore new and uncharted territories of expression.
By recognising the significance of human skill and judgement in determining originality, as established in the Eastern Book Company case, we can ensure that AI-generated works are not mere imitations or replications of existing creations but rather valuable contributions to the realm of human creativity and intellectual property.
International perspective
Globally, the legal response to AI and copyright infringement varies. In most jurisdictions, fair use is used as a principal to decide if input used for training is covered under copyright infringement or not. The status of some countries is as follows:
In the European Union, the Copyright Directive, 2019 highlights the need for balancing innovation with intellectual property rights protection.
In the United States, the fair use doctrine has played the major role in dealing with the issue of generative AI and copyright infringement, which is clear from cases like Authors Guild vs. Google, Inc. and Warhol vs. Goldsmith. Though companies like OpenAI and Google have started the practice of entering into licensing agreements with companies like Reddit for training purposes, this highlights that the fair use doctrine may not be enough to protect such material from copyright infringement in the future when more such cases arise.
In France, the action of the Competition Authority charging Google $250M for using news articles without proper licensing to train Gemini shows their position in such cases, i.e., they do not use fair use doctrine and mandate licensing for such actions.
In Japan, it is allowed to use copyrighted work for the purpose of training data by generative AI.
In India, the Copyright Act of 1957 and its subsequent amendments serve as the primary legal framework for addressing issues related to copyright protection and infringement. However, at present, there is no specific law that solely deals with the issue of copyright infringement in the digital domain. This is because when the Copyright Act was enacted in 1957, the concept of digital technology and the internet was not prevalent. As a result, the existing provisions of the act were drafted with a focus on traditional forms of copyright infringement, such as unauthorised reproduction, distribution, and public performance of copyrighted works in physical formats.
Nevertheless, with the rapid advancement of technology and the widespread use of the internet, digital copyright infringement has emerged as a significant concern. Digital technologies have made it easier for individuals to reproduce, distribute, and share copyrighted content without authorisation, raising questions about the adequacy of the existing legal framework to address these challenges.
Recognising the need for a more comprehensive approach to addressing digital copyright infringement, there have been calls for the enactment of specific legislation or amendments to the existing Copyright Act to incorporate provisions that specifically address the unique issues posed by the digital environment. Such legislation could include measures to enhance enforcement mechanisms, provide for effective remedies for copyright holders, and address challenges related to online intermediaries and user-generated content.
While the current legal framework in India may not explicitly address digital copyright infringement, courts have attempted to interpret the existing provisions of the Copyright Act in light of technological developments. In some cases, courts have applied traditional copyright principles to digital contexts, holding that unauthorised copying, distribution, or public performance of copyrighted works in digital formats constitutes infringement. However, the absence of specific provisions tailored to digital copyright infringement can sometimes lead to uncertainties and challenges in enforcing copyright rights in the digital realm.
As the legal landscape evolves, it is possible that specific regulations or amendments to the Copyright Act may be introduced to address digital copyright infringement more comprehensively. Such developments would aim to provide a more robust legal framework that is better equipped to protect the rights of copyright holders and promote creativity and innovation in the digital age.
Case laws related to concept of fair use
In Andy Warhol Foundation for the Visual Arts, Inc. vs. Lynn Goldsmith, Warhol created artworks based on Goldsmith’s 1981 photograph of the Prince, which was licensed by Vanity Fair in 1984. After Warhol’s death, in 2016, his foundation licensed his work to Condé Nast. Due to this event, Goldsmith filed a copyright infringement suit against the foundation, to which they argued that there was no infringement according to Section 107 of the US Copyright Act, which deals with the doctrine of fair use.
After careful analysis of the case, the court established that Warhol used significant parts of Goldsmith’s photograph that could substitute Goldsmith’s work in the market, establishing the fact that Warhol’s work was not transformative enough to be constituted as “original.” On the basis of the above observation, it was held that Warhol’s use did not qualify as fair use. This decision highlights the need for creativity and originality in the work of an author to use the doctrine of fair use as an exemption in cases related to generative AI and copyright infringement.
In Civic Chandran vs. C. Ammini Amma, the Kerala High Court addressed the issue of copyright infringement involving a theatrical performance. Chandran used a novel by Amma illegally without any licensing. The court ruled against Chandran and stated that Chandran’s use was not transformative enough and thus the fair use defence can’t be claimed.
This case highlights the importance of obtaining proper authorisation for using someone’s copyrighted work. Apart from that, it was held that the Hon’ble Court may allow the “fair use” doctrine’s four-factor test established in the US in such cases because it is similar to “fair dealing” under Section 52 of the Copyright Act, 1957.
Tips Industries Ltd. vs. Wynk Music Ltd., a landmark case in the realm of copyright law, stands as a testament to the significance of obtaining proper licenses for digital music streaming. At the heart of this case lies Wynk Music Ltd.’s unauthorised use of Tips Industries Ltd.’s music catalogue on its streaming platform. This blatant disregard for copyright laws prompted Tips Industries Ltd. to seek legal recourse.
The court, recognising the gravity of the matter, delved into the intricacy of copyright laws and their applicability to digital music streaming. It emphasised that under the Copyright Act, streaming platforms are obligated to secure licenses from copyright holders before making music available to their users. Wynk Music Ltd.’s failure to comply with this legal requirement constituted a clear case of copyright infringement.
The court’s ruling in favour of Tips Industries Ltd. sent a strong message to the music industry, highlighting the importance of respecting intellectual property rights. It established a precedent that digital music streaming platforms cannot operate outside the boundaries of copyright laws and must obtain the necessary licenses to legally distribute music content.
This case serves as a reminder that copyright protection extends beyond traditional forms of media and encompasses the digital realm as well. It underscores the need for music streaming platforms to engage in ethical practices and ensure that they have the proper permissions in place to avoid copyright infringement.
Furthermore, the Tips Industries Ltd. vs. Wynk Music Ltd. case brought to light the growing complexities surrounding digital music streaming rights. It sparked discussions about the need to modernize copyright laws to keep pace with technological advancements and evolving business models in the music industry.
Drawbacks of using copyrighted content for AI training
Generative AI plays an important role in the development of all sectors of the economy but it comes along with few legal and ethical concerns, such as adverse effects on confidentiality, loss of control by original authors’ over their copyrighted work, undue market advantage by large tech firms, biassed training data leading to biassed and generalised outputs, undue legal advantage taken by licensing companies, disadvantage for poor due to huge requirement of funds, lack of creativity and specialisation in delivered content, job destruction along with job augmentation (particularly harmful for clerical roles) and many more.
Way forward
To deal with the complex issue of generative AI and copyright infringement in a proper way, legislative updates are required. There is a need to lay down proper legal, ethical, technological, and industrial standards along with international harmonization. Balancing innovation through AI with the intellectual property rights of original authors is a must. To achieve harmony amongst both, there is a need to establish proper guidelines that AI developers will be obligated to follow, thus protecting the rights of both the AI developer and the original owner.
Indian regulations on AI copyright infringement
The Indian Copyright Act, 1957, and the Patents Act, 1970 are the two primary pieces of legislation governing intellectual property rights in India. However, neither of these acts has any provisions that specifically address the issue of copyright infringement by artificial intelligence (AI). This is because AI was not a significant factor when these acts were drafted.
The Copyright Act, 1957, defines copyright as “the exclusive right to do or authorise certain acts in relation to a work.” These acts include the reproduction, distribution, adaptation, and public performance of a work. The act also states that copyright protection exists in original literary, dramatic, musical, and artistic works. AI-generated works may fall within these categories, but the act does not explicitly address this issue.
The Patents Act, 1970, defines a patent as “an exclusive right granted for an invention.” An invention is defined as “a new product or process involving an inventive step and capable of industrial application.” AI-generated inventions may meet these criteria, but the act does not explicitly address this issue.
In addition to the Copyright Act and the Patents Act, there are a number of other laws that may be relevant to AI-generated content. These include the Trademarks Act, 1999, the Designs Act, 2000, and the Information Technology Act, 2000. However, none of these laws specifically address the issue of AI-generated content.
The lack of legal provisions governing AI-generated content creates a number of challenges. For example, it is unclear who owns the copyright to an AI-generated work. Is it the AI itself, the person who created the AI, or the person who used the AI to create the work? It is also unclear whether AI-generated works are eligible for patent protection.
The lack of legal clarity surrounding AI-generated content is a significant barrier to the development and use of AI. It is important for lawmakers to address this issue in order to provide legal certainty for businesses and individuals who are using AI.
Until the law is clarified, businesses and individuals who are using AI should take steps to protect their intellectual property rights. This may include registering copyrights and patents for AI-generated works and entering into agreements with AI developers that specify who owns the intellectual property rights to AI-generated content.
Many steps have been taken forward to promote AI development along with copyright protection. In 2018, NITI Aayog introduced an AI strategy focussing on responsible development by following ethical concerns and imparting public education on these subjects. After this, in 2022, the Centre of Excellence in AI was established, which launched the IndiaAI portal and joined the Global Partnership on AI.
Till now, no AI-specific laws have been established but the Digital Personal Data Protection (DPDP) Act, 2023 may apply to AI-related cases to some extent. It neither bans scrapping of publicly available data without permission nor allows it. Without specific guidelines, few AI companies might take this silence as implied consent, which might ultimately lead to privacy and ethical concerns. This void is really confusing and it risks widespread data collection without proper permission, which might ultimately lead to copyright infringement. Therefore this Act isn’t very helpful and there is a stagnant need to develop comprehensive laws that explicitly deal with generative AI-related issues.
Whereas, in response to questions related to the need for new laws dealing with generative AI and copyright infringement issues, Union Minister Shri. Som Parkash stated that the current IPR regime is sufficient to deal with issues related to infringement and ownership for both human and AI-generated content and there is no need for any amendments in existing IP laws or for the creation of new laws.
Therefore, to keep up with technological advancements and to deal with AI and copyright infringement issues, India should mandate proper licensing and an explicit consent policy for using data as training data by AI platforms. Apart from that, there is a need for detailed fair use guidelines to ensure transparency and to promote ethical standards. India should harmonise its practices with international laws and should keep on updating its policies with time in accordance with the changes and requirements of the surrounding environment.
International perspective on AI copyright infringement
The U.S. has taken several steps towards AI regulation. In August 2019, NIST released a report on AI standards and risk management regulations. In April 2021, FTC, EEOC, CFPB, and DOJ published a joint statement stating that existing laws apply to AI in the U.S. Then in May 2021 and January 2023, NITA and NIST, respectively, released their reports on AI Accountability and AI Risk Management guidelines. In October 2023, even President Biden issued an order guiding responsible AI development. Thus, the U.S. is dealing with this issue in a proper legal and ethical manner.
Japan issued the 2019 Report, “Social Principles of Human-Centric AI.” According to this report, Japan promoted a human-centric approach for AI development.
China highlighted the importance of AI by incurring huge investments in the AI sector in its 13th and 14th five-year plans. Apart from that, the Cyberspace Administration of China introduces proper rules for generative AI in July, 2023.
The UK hosted the first global AI summit in November 2023 at Bletchley Park. Therefore, in every part of the world, countries are taking major steps towards AI development without affecting anyone’s rights.
Conclusion
The merger of generative AI and copyright infringement is both challenging as well as rewarding. With advancements in technology, there is a dire need to resolve the conflict between copyright infringement and AI development. There is a vital need to deal with the issues of infringement, originality, ownership, non-human IP rights, and many more. Almost every part of the world is dealing with this issue in its own way. Some countries have already developed a practice to deal with it, while others are still in the process of choosing what’s best for their country. There are a few landmark cases, such as Authors Guild vs. Google and Andy Warhol Foundation vs. Lynn Goldsmith, which highlight the evolving legal framework in relation to AI copyright infringement issues by primarily focussing on fair use doctrine as a defence.
Future policies need to strike a balance between AI development and copyright protection. They need to protect both innovation and existing legal and ethical rights. Creativity and originality are considered basic fundamentals to decide whether content is copied or is an original piece. International harmony must be maintained while developing such laws. There is a need to develop AI for a better future, and so is the need to protect existing IP rights so that the drive for innovation among creators stays unhampered.
The rapid expansion of AI technology by tech leaders like Elon Musk and Sam Altman is highly unregulated. The present government is not able to manage both innovation and risks together. There are no proper legal provisions available anywhere to deal with such an important issue. India, especially, is way behind in this arena. To reach great heights in the future, it is very important for AI developers, users, authors, and policymakers to address these issues properly and reach an amicable solution for both segments of this area, i.e., copyright holders and AI developers. There is a crucial need to support creativity and innovation while upholding copyright laws.
References
- https://blog.ipleaders.in/copyright-infringement-in-the-digital-age/#:~:text=A%20legal%20letter%20sent%20by,as%20a%20copyright%20infringement%20notice.
- https://pwonlyias.com/current-affairs/generative-ai-and-copyright-issues/#:~:text=The%20use%20of%20copyrighted%20materials,generated%20by%20the%20AI%20concerned.
- https://pwonlyias.com/current-affairs/ilo-report-on-generative-ai/
- https://research.aimultiple.com/generative-ai-copyright/
- https://crsreports.congress.gov/product/pdf/LSB/LSB10922
- https://hbr.org/2023/04/generative-ai-has-an-intellectual-property-problem
- https://www.ijlt.in/post/balancing-indian-copyright-law-with-ai-generated-content-the-significant-human-input-approach
- https://guides.lib.usf.edu/c.php?g=1315087&p=9690822
- https://scholar.smu.edu/cgi/viewcontent.cgi?article=1360&context=scitech
- https://www.linkedin.com/pulse/training-ai-generate-content-violation-copyright-laws-sonisvision-n2kre/
- https://www.csis.org/blogs/perspectives-innovation/informing-innovation-policy-debate-key-concepts-copyright-laws
- https://www.cmswire.com/digital-experience/ai-copyright-infringement-quandary-generative-ai-on-trial/
- https://www.medianama.com/2024/02/223-commerce-ministry-copyright-laws-gen-ai/
- https://www.mondaq.com/india/copyright/1348418/legal-implications-of-ai-created-works-in-india
- https://blog.ipleaders.in/critical-analysis-of-ai-and-copyright-protection/
- https://blog.ipleaders.in/artificial-intelligence-and-laws-persisting-over-it/
- https://blog.ipleaders.in/understanding-copyright-issues-entailing-deepfakes-in-india/
- https://spicyip.com/2024/02/spicyip-tidbit-union-minister-of-state-for-commerce-and-industry-clarifies-current-ipr-regime-sufficient-for-ai-works-protection.html
- https://spicyip.com/2024/04/training-genai-infringement-or-fair-use.html
- https://harvardlawreview.org/print/vol-137/andy-warhol-foundation-for-visual-arts-inc-v-goldsmith/
- https://www.supremecourt.gov/opinions/22pdf/21-869_87ad.pdf
- https://pib.gov.in/PressReleasePage.aspx?PRID=2004715
- https://indiankanoon.org/doc/1062099/
- https://www.nist.gov/system/files/documents/2019/08/10/ai_standards_fedengagement_plan_9aug2019.pdf https://www.ntia.gov/issues/artificial-intelligence/stakeholder-engagement/request-for-comments
- https://www.ftc.gov/system/files/ftc_gov/pdf/EEOC-CRT-FTC-CFPB-AI-Joint-Statement%28final%29.pdf
- https://www.scconline.com/
![](https://blog.ipleaders.in/wp-content/uploads/2021/01/Internship-Banner.jpg)