The Legal Landscape Of Copyright In The Era Of Artificial Intelligence
Author: LawSphereX
Category: Technology, Cyber & Privacy
Published:
Updated:
THE LEGAL LANDSCAPE OF COPYRIGHT IN THE ERA OF
ARTIFICIAL INTELLIGENCE
Monikornika Dutta, Lecturer, Apex Professional University, Pasighat, Arunachal Pradesh
ABSTRACT
AI is revolutionizing industries through advanced data analysis and automation capabilities. However, these advancements have also given rise to legal challenges, particularly concerning AI’s ability to collect, process, store, and analyze vast amounts of data. Since data is fundamental to AI, its use raises significant legal concerns, especially when AI generates new content based on pre-existing data, potentially infringing on copyright laws. During the training of AI model, large datasets are used, which sometimes includes copyrighted material. The widespread availability of such data on the internet makes it easily accessible for training purposes. Generative AI can produce content that closely resembles copyrighted works, leads to legal ambiguity.
As AI itself lacks legal recognition, determining liability for such copyright infringements remains a complex issue for the legal community. This research aims to critically analyze AI-related copyright violations and identify potential legal uncertainties and challenges in copyright enforcement related to AI while providing recommendations for future policymaking. Keywords: Artificial Intelligence; Copyright Infringement; Generative AI; Data; Legal Liability
1. Introduction
The advancement of AI has revolutionized various industries by harnessing advanced data analysis and automated processes. It has the ability to collect, process, store, and analyze vast amounts of data raising significant legal challenges of copyright violation. These datasets forming the backbone of AI, mostly contain copyrighted material. AI models are trained on the data that is already in the public domain (internet). This accessibility of data through the internet further facilitates the AI tools to generate novel content. With the rise of Generative AI models like GPT-4, Midjourney, and Stability AI trained to create new content based on existing data on command of the user is currently questioning the existing Intellectual Property Rights. Moreover, the developers of the AI model don’t properly cite the copyright holders when copyrighted data are used in developing an AI model.
This raises questions about the legality of the use of copyrighted material for training purposes of AI. When the model is trained on existing copyrighted material then the output generated by it may include that data provided by the AI model is likely to infringe upon the copyright holder. The close resemblance of AI-generated content to that of copyrighted material creates new legal ambiguities such as the legal status of AI. As AI itself lacks global recognition, establishing legality for infringement of copyright presents a complex challenge for the legal community. This research paper aims to explore the intersection of AI technology and Copyright violation, the implications of AI-generated content on IPR protection, and the need for emerging legal frameworks to deal with challenges posed by these technologies.
By analyzing the current legal frameworks around the world such as the US, EU, China, and India, this paper aims to add to the ongoing discussion for balancing innovation with the rights of the content creator and copyright holders in the AI-driven world.
1.2 Research Problem and Objectives
AI’s reliance on vast datasets raises concerns about copyright infringement. As the models are trained based on data available on the internet, they can include original copyrighted work. Most of the time, these data are taken without the permission of the copyright holder. This
becomes a topic of concern when AI-generated content resembles existing copyrighted work. The absence of legal recognition of AI further complicates liability determination. This research aims to:
1. Analyze AI-related copyright violation and their implications
2. Identifying Gaps and uncertainties in existing copyright laws
3. Exploring international and national legal frameworks addressing AI and Copyright
4. Propose legal and policy recommendations for mitigating AI-related copyright issues.
1.3 Research Methodology
This research adopts a doctrinal research approach, analyzing legal texts, case laws, judicial precedents and policy documents. A comparative analysis of AI-related copyright disputes across different jurisdictions will provide insight into global legal trends and enforcement challenges. The research relies heavily on secondary sources existing literature and online resources.
2. The Intersection of Artificial Intelligence and Copyright
The rapid advancement of Artificial Intelligence has transformed the way creative works are produced. It raises complex questions about authorship, ownership, and copyright protection in the era of AI. Traditionally, copyright law is crafted to protect works created by human authors, but with the development, the role of AI in generating artistic, literary, and musical works challenges these already established principles of law. Generative AI is a type of AI that can create new content and ideas, such as conversations, stories, images, videos, and music. For example, AI is capable of learning from English vocabulary and crafting a poem by using the words it has processed. Organizations can use Generative AI for numerous purposes such as the development of chatbots, content creation, and product innovation and design.
2.1 AI and Copyright Infringement
Both AI and Generative AI raise copyright issues regarding infringement and authorship rights. Generative AI tools like Midjourney, OpenAI’s DELL-E and Stability AI are trained on vast
amounts of data which has already been copyrighted. While training such models, the company doesn’t always get permission from the copyright holder. Instead, they claim that the copyrighted data are used for fair use, which becomes an exception on their side. The use of copyrighted material (data) without permission of the owner for training of the AI model can be considered unauthorized use. These models learns from the data already copyrighted and give results that sometimes resemble the original copyrighted work. Such kind of output raises concern about derivative work. Along with this, the AI doesn’t properly cite or give credit to the original creators, which can be considered a violation of moral rights.
2.2 Training of AI models with copyrighted data
In Getty Images v. Stability AI, 20231, a UK-based lawsuit where Getty Image sued Stability AI, the developer of Stable Diffusion (an AI model) for scraping millions of copyrighted images, without the consent or authorization to train its AI model. The lawsuit also accuses Stability AI as some of the outputs of Stable Diffusion contain traces of Getty’s watermark, which shows that the AI model was trained with copyrighted material which can potentially mislead the consumers and devalue the copyrighted material. Stability AI in its defense argued on the basis of the Fair Use doctrine but unfortunately, UK law doesn’t recognize this principle under their law as the US. Also, Stability AI stated that it did not directly scrape images from Getty, rather it used images that are already available on the internet.
The case is still pending before the court but have the potential to deliver a landmark Judgment for copyright protection and AI training. If Getty wins the case it could force AI companies to license images before training the model and if Judgement comes in favour of Stability than the companies will have freedom to use copyrighted material for training the AI model.
2.3 EU’s Stand on Regulating AI
The EU’s Artificial Intelligence Act, of 2024 is the first act to govern and regulate AI. It aims to establish a regulatory framework for AI. Like the EU’s General Data Protection Regulation (GDPR) in 2018, the Artificial Intelligence Act has become a global standard for regulating AI. This act classifies AI systems on a risk-based approach so as to impose compliance and ensure safety and security of individual’s right. Although the act doesn’t directly address copyright 1Getty Images & Others v. Stability AI, [2025] EWHC (Judiciary UK) (Jan. 14, 2025), https://www.judiciary.uk/wp-content/uploads/2025/01/Getty-Images-and-others-v-Stability-AI-14.01.25.pdf.
issues, but it does requires the discloser of information about the datasets used to train AI models by AI providers and also requires specific authorization from copyright owners for the use of their work to train Generative AI model. Along with this the act also mandates for adhering to the EU’s copyright law once AI system developed outside EU enters into European market. In the case French Publisher and authors vs Meta 20252 the petitioner accused Meta a social media company for using their copyrighted work without permission for training its AI model. Meta has used several copyrighted books, articles, and other literary works for training datasets without authorization. Also Meta was earning profits from these AI models without compensating the original authors and have not yet disclosed information about what kind of work they have used in training the model.
This case is now in early stage with legal proceedings to unfold throughout 2025. The EU’s AI Act provides for Transparency requirement under which the providers of Generative AI models must disclose when the content is AI generated and publish summaries of data used for training the model. 3 This lawsuit highlights the ongoing debate of technological advancement and the protection of creative work, a balance that the AI Act seeks to address through its regulatory framework.
3. Copyright Status of AI-generated and AI-assisted works in Different Jurisdictions
AI involvement in creative work can be classified into two categories, AI-generated and AIassisted works.4 AI-generated works are autonomously created by AI systems without human intervention. These works often result from machine learning algorithms analyzing vast datasets and producing outputs that may not have been anticipated by human programmers.5 So, in AI-generated work, the authorship cannot be granted.6 But in the case of AI-assisted works there is a significant involvement of human input, where AI serves as a tool to enhance 2 French Publishers and Authors Sue Meta over Copyright Works Used in AI Training, Euronews (Mar. 13, 2025),https://www.euronews.com/culture/2025/03/13/french-publishers-and-authors-sue-meta-over-copyrightworks-used-in-ai-training.
3 European Parliament, EU AI Act: First Regulation on Artificial Intelligence (June 1, 2023),
https://www.europarl.europa.eu/topics/en/article/20230601STO93804/eu-ai-act-first-regulation-on-artificialintelligence. 4 World Intellectual Property Organization, WIPO IP and AI Meeting, WIPO/IP/AI/2/GE/20/1 Rev., para. 12 (2020), https://www.wipo.int/edocs/mdocs/mdocs/en/wipo_ip_ai_2_ge_20/wipo_ip_ai_2_ge_20_1_rev.pdf.
5 Artificial Intelligence and Copyright, Vanderbilt Univ. Faculty Publs.,
https://scholarship.law.vanderbilt.edu/cgi/viewcontent.cgi?article=2176&context=faculty-publications (last visited, Mar. 19, 2025).
6 V.K. Ahuja, ARTIFICIAL INTELLIGENCE AND COPYRIGHT: ISSUES AND CHALLENGES, ILI Law
Review 5(2020), https://ili.ac.in/pdf/vka.pdf.
or facilitate the creative process while the human retains primary control over the final output. So in AI-assisted Work, authorship can be granted.7
3.1 United States
The U. S. Copyright Act protects the original work of authorship, which means that copyright can be granted to the content created by a human being. The U. S. Copyright Office has rejected many applications for copyright, stating a lack of human intervention. The US 9th Circuit Court of Appeal in the case of Naruto v. Slater (2018)8 famously known as ‘Monkey Selfie Copyright Dispute’ held that non-humans cannot be a copyright holder of Artistic Work. 9 Monkey Selfie Copyright Dispute Source: https://en. wikipedia. org/wiki/Monkey_selfie_copyright_dispute The background of the case is that the monkey named Naruto clicked several selfies on the camera of David Slater (Photographer) who left his camera in the junglee unattended. One of the selfies taken by the monkey became widely circulated, which led to the dispute over ownership of the copyright. The U. S.
Copyright Office refused to grant Copyright to Naruto as the Act doesn’t provide provision for it. The significance of this case is that if the selfie taken by a monkey cannot receive Copyright, then the text, image, and music solely generated by AI without human intervention also be denied copyright. But some of the works originated by AI is copyrightable in U. S subject to the condition that the content must pass the test of de
7 Id. at 6.
8 Monkey Selfie Copyright Dispute, Wikipedia, https://en.wikipedia.org/wiki/Monkey_selfie_copyright_dispute (last visited Mar. 19, 2025).
9 Artificial Intelligence and Copyright, Wikipedia,
https://en.wikipedia.org/wiki/Artificial_intelligence_and_copyright#cite_note-crs-4 (last visited Mar. 19, 2025).
minimus10 which stands for minimum AI interference. If the creator has used limited AI resources in creation of an end product than he can claim Copyright under U. S law. This difference of approach for granting of Copyright to AI was seen when Jason Matthew Allen in Sep, 2022 went to the Copyright Office for Registration of his work (image) which he created using Midjourney and other upscaling tools. The image has won many competition and has become the first AI-generated Image to won reputed awards. Théâtre D’Opéra Spatial Source: https://en. wikipedia. org/wiki/Th%C3%A9%C3%A2tre_D%27op%C3%A9ra_Spatial In 2022 U. S Copyright Office initially refused to grant copyright to the whole image as part of it were AI generated and contribution to the image by Allen and Midjourney are “inextricable merged, inseparable contribution”.
The Copyright Office argued that the copyright can be granted to the parts of the art edited through Photoshop, but the portion of the image created via Midjourney and Gigapixel AI had to be excluded. Allen filed second request on July,2023 for reconsideration. On Sep,2023 The Copyright Review Board made final decision and found that Théâtre D'opéra Spatial 11 was not eligible for Copyright Protection under U. S Law as human creative input was de minimis, and AI generated elements dominated most parts of the image. In January 2025 the US copyright office clarified that AI assisted work where human creativity remains evident can be copyrighted,12 which includes creative adaptation of prompts for AI
10 Id. at 9
11 Théâtre D’Opéra Spatial, Wikipedia,
https://en.wikipedia.org/wiki/Th%C3%A9%C3%A2tre_D%27op%C3%A9ra_Spatial (last visited Mar. 19, 2025). 12 AI and Copyright Office Debate: Artificial Intelligence and Rights, AP News, https://apnews.com/article/aicopyright-office-artificial-intelligence-363f1c537eb86b624bf5e81bed70d459 (last visited Mar. 19, 2025).
generators or usages of AI to assist in the creation of filmmaking.13
3.2 European Union
The European Union is actively addressing the complex relationship between AI and copyright law. It aims to balance technological advancement with the protection of Intellectual Property rights. Currently, EU doesn’t provide copyright protection to AI-generated works because the current legal framework of copyright in the EU is based on the principle of human creativity and intellectual level. The copyright law of the EU also requires human authorship to get copyright protection. The Court of Justice of the European Union recently held that copyright only applies to author’s own intellectual creation. 14 As AI lacks human intent and consciousness, AI-generated work cannot be granted a copyright. The Municipal Court in Prague in a recent case in the Czech Republic15 determined that a graphic generated by AI without significant human input cannot be protected under Copyright.
This highlights the necessity of human creativity for copyright eligibility.
3.3 China
The Beijing Internet Court, in Li v Liu16, has granted Copyright to an AI-generated Image Spring Breeze Brings Tenderness for the first time in 2023. The court passed the judgment after determining that the work has met the originality requirement of the Chinese Copyright Law.17
13 Artificial Intelligence and Copyright, Wikipedia,
https://en.wikipedia.org/wiki/Artificial_intelligence_and_copyright#cite_note-14 (last visited Mar. 19, 2025).
14 AI and Copyright: First Ruling in European Court, Novagraaf (2024),
https://www.novagraaf.com/en/insights/ai-and-copyright-first-ruling-european-court (last visited Mar. 19, 2025).
15 No Copyright for AI Creations in the Czech Republic, LegalAir Blog (Aug. 12, 2024),
https://legalair.nl/blog/2024/08/12/no-copyright-for-ai-creations-in-the-czech-republic-what-does-this-mean-forthe-rest-of-europe/. 16 Yaxuan Wang & Jing Zhang, Beijing Internet Court Grants Copyright to AI-Generated Image for the First Time, Kluwer Copyright Blog (Feb. 2, 2024), https://copyrightblog.kluweriplaw.com/2024/02/02/beijinginternet-court-grants-copyright-to-ai-generated-image-for-the-first-time/.
17 Beijing Internet Court Recognizes AI Copyright, China IP L. Update (Jan. 2024),
https://www.chinaiplawupdate.com/2024/01/beijing-internet-court-releases-translation-of-li-vs-liu-recognizingcopyright-in-generative-ai/.
Spring Breeze Brings Tenderness Source:https://www. kwm. com/cn/en/insights/latest-thinking/china-s-first-case-oncopyrightability-of-ai-generated-picture. html This judgment implies that with sufficient human intellectual effort, AI-generated content can be copyrightable. In Lin Chen v. Changshu Qin Hong Real Estate Development Co. , Ltd. and Hangzhou Gauss Air Film Technology Co. , Ltd. (2024), the Changshu Municipal People’s Court recognized the copyrightability of AI-generated digital 2D artwork named ‘Companion Heart’ created by artist Lin Chen using Midjourney and Photoshop. Companion Heart Source: https://michael7924. substack. com/p/chinas-second-local-court-recognizes The defendant was found to have infringed the right of the petitioner by using the image for their promotional work without authorization and also creating a 3D version of that image.
The court held that Lin’s selection of prompts using and modification of image through AI image modification and processing software shows and reflects his unique selection and arrangement of prompts, hence, the image can be considered as original and belongs to protection under
copyright law.18 The court also ruled that the defendant had violated the copyright of the plaintiff by disseminating the image on the internet without permission and ordered the defendant to issue a public apology and compensate the plaintiff with RMB 10,000 for infringement.19
3.4 India
Looking at the current legal framework of India, copyright cannot be granted to AI-generated work. Under the Indian Copyright Act,1957, only works created by human authorship are entitled to copyright protection, which directly excludes AI-generated content from copyright protection. The Ministry of Commerce and Industry has clearly stated that the existing IPR laws in India are adequately capable of addressing AI-related copyright issues and also suggested that there is no immediate need for new legislation. 20 In India the issues related to Copyright and AI is different to that of other countries. Till now no cases were registered for claiming copyright for AI-generated content. However, Indian Courts were dealing with the issues of training AI models with Copyrighted content.
In 2024 ANI (Asia News International) filed a copyright infringement lawsuit against OpenAI for using its copyrighted material to train AI models like ChatGPT without authorization in the Delhi High Court. 21 ANI argued that : I.
Not all of their contents are publicly accessible but only to the subscribers. Hence OpenAI doesn’t have license to copy, store and use such material to train the AI model.22 II.
ChatGPT was falsely generating news articles attributing ANI which could damage ANI’s credibility. III.
OpenAI is generating economic benefits from AI-generated news summaries without
18 Chinese Court Rules AI Images Eligible for Copyright, National Law Review (2024),
https://natlawreview.com/article/chinese-court-again-rules-ai-generated-images-are-eligible-copyrightprotection.
19 Id at 18.
20 Press Info. Bureau, Government’s Stance on AI Copyright, PIB India (2025),
https://pib.gov.in/PressReleasePage.aspx?PRID=2004715.
21 ANI vs OpenAI Legal Case, Law Asia (2025), https://law.asia/ani-vs-openai-legal-case/.
22 ANI vs OpenAI Legal Case, Law Asia (2025), https://law.asia/ani-vs-openai-legal-case/.
paying or crediting (citing) the original creators. OpenAI in its defense argued that: I.
AI models learn from publicly available data and do not reproduce copyrighted work verbatim. II.
OpenAI argued that as a US based company it is not subject to Indian Courts as there servers were stored outside India, no physical presence, and the models of AI were trained in the US. III.
ChatGPT does not reproduces exact content to that of ANI rather it produces results based on training of data. Also ANI doesn’t show any evidence of specific instance of Infringement. Delhi High Court asserted Jurisdiction and allowed the case to proceed. The court also appoints an amicus curiae to assist in the legal proceedings of the case to ensure that the case is considered from all relevant perspectives. Other reputed Indian media houses are also interested in joining the case because of the rising concern about AI’s impact on Journalism. The case is ongoing and has the potential to set legal precedents for AI regulation, training, and IPR protection in terms of copyright.23
4. Granting of Copyright Protection to the AI-generated Work
The current legal jurisprudence of copyright does not clearly provide provisions for the copyrightability of AI-generated work. The Copyright laws in force now have been shaped and established through several international conventions, treaties, and agreements that provide global standards for copyright protection. These instruments clearly define who can own copyright, what amounts to copyright infringements, and other legal provisions to deal with issues of copyright. The Berne Convention is only applicable to human-created works, which directly makes AI-generated work ineligible for copyright protection. Under TRIPs agreement also same status has been maintained as it complies with the Berne Convention only. Under the WIPO Copyright Treaty and The WIPO Performances and Phonograms Treaty a similar
23 Id at 21
position has been maintained. 24 These show that at the international level all these major treaties have recognized only the human creators as capable of holding the copyright. Hence AI-generated works remain outside the formal copyright jurisprudence unless a human demonstrates creative control over the AI output. However, these International copyright frameworks lack clear guidelines for AI-generated content. The Judiciary and Legislature in different jurisdictions around the world are currently filling these gaps through judgments and Policy implementation regarding the Copyrightability of AI-generated works. In 2017, AIVA (Artificial Intelligence Virtual Artist), an AI-generated music composing tool was granted copyright by the French and Luxembourg author’s right society, SACEM for its composition.
25 The grounds for granting Copyright to AIVA is that its compositions demonstrate the level of creativity and originality to that of a human composers. With this recognition AIVA has become the first ever AI to hold copyright and receive royalties for the music it composes. 26 Saudi Arabia has also granted citizenship to the Humanoid AI robot, Sophia, in 2017. Dr. David Hanson who created Sophia, in one of his paper titled “Entering the Stage of Living Intelligence and Android Society”, writes that AI robots are going to claim IPR protection for their creation. 27 Different Jurisdictions have interpreted the issue of authorship differently, which means that granting copyright to AIVA or legal status to Sophia may not be universally accepted. However, both these cases have highlighted the growing legal debate around AI and Copyright.
While some jurisdictions have recognised AI-generated content under certain conditions, most copyright laws still require human authorship.
5. Determining the Liability for AI-created Work
With the development and widespread use of AI-generated content, the existing legal
24 V.K. Ahuja, ARTIFICIAL INTELLIGENCE AND COPYRIGHT: ISSUES AND CHALLENGES, ILI Law
Review 281 (2020), https://ili.ac.in/pdf/vka.pdf.
25 V.K. Ahuja, ARTIFICIAL INTELLIGENCE AND COPYRIGHT: ISSUES AND CHALLENGES, ILI Law
Review 276 (2020), https://ili.ac.in/pdf/vka.pdf. 26 AIVA Technology: Composing Music Using AI, Harv. Dig. Platform, https://d3.harvard.edu/platformdigit/submission/aiva-technology-composing-music-using-ai/ (last visited Mar. 19, 2025).
27 Robots and Civil Rights: AI’s Future, Independent (May 24, 2018),
https://www.independent.co.uk/tech/robots-civil-rights-android-artificial-intelligence-2045-destroy-humanssophia-singularity-a8367331.html
frameworks struggle to determine the liability factor for AI. As the legal status has not been granted to AI, the legal system struggles to determine liability in case of AI-generated copyright infringement. In such a scenario, the liability can fall on different parties depending on how AI is controlled and used. Possible Liable Parties can be- I.
AI Developers- Companies that develop AI models, such as OpenAI, Meta or Stability AI, could be held liable if their models are trained on copyrighted material without permission, which amounts to copyright infringement. Developers in such a situation may face a lawsuit. For example: Getty Image v. Stability AI, 2023 & French Publisher and authors v. Meta, 2025. II.
AI User- The user can be the one using the output of AI. If someone publishes the AIgenerated content that closely resembles copyrighted work or spreads misinformation, then the users can be held liable. For example, the case of Sarah Anderson v. Stability AI 2023. 28 The plaintiff has filed a lawsuit against Stability AI, Midjourney, and DeviantArt-alleging that these companies infringed the copyright of the plaintiff’s artwork already registered with the U. S. Copyright Office, to train AI models without the consent of the copyright holder. The plaintiff claimed that Stability AI’s Stable Diffusion model was trained on the copyrighted works without authorization, resulting in derivative works that infringed the plaintiff’s rights. Stability AI filed a motion to dismiss the claims.
However, the court denied this motion regarding the direct infringement claims, stating that the plaintiff’s allegations were sufficient to proceed. As of now, the case is ongoing. Several defenses also available for AI Users and Developers to escape from liability such as the principle of Fair use, lack of intent or use of AI content in good faith etc. Despite these defenses, the absence of well defined legal framework for determining the liability of AIgenerated copyright infringement complicates the determination of liability. Courts were
28 Zach Schor, Andersen v. Stability AI: The Landmark Case Unpacking the Copyright Risks of AI Image
Generators, JIPEL, NYU Journal of Intellectual Property and Entertainment Law Dec 2, 2024 https://jipel.law.nyu.edu/andersen-v-stability-ai-the-landmark-case-unpacking-the-copyright-risks-of-ai-imagegenerators/#:~:text=The%20induced%20infringement%20claim%20against,at%20the%20summary%20judgme nt%20stage.
currently treating the AI as a tool, placing the liability on human users and developers. But there is also a technical complexity involves which is called the AI’s Black Box Theory. Sometimes these Generative AI models give results in a way that cannot be understood by the developers of the model and such results cannot be traced due to the opacity of Black Box Theory. As AI-generated content is growing, new legal frameworks establishing distinct liability for AI to deal with these new forms of copyright infringement issues are a need of the hour.
6. Conclusion
This research has examined the legal complexities surrounding AI-generated content and copyright law. The intersection of AI and Copyright law shows several ethical, legal, and regulatory challenges. This highlights the growing tension between technological innovation and Intellectual Property Protection. With the advancement of AI, its potential for generating music, photographs, text, or other creative works has raised the fundamental question of liability, authorship, and ownership under the existing legal framework of copyright protection. While international conventions such as the Berne Convention, TRIPs Agreement, and WIPO treaties emphasized on human authorship, cases like AIVA’s recognition in Luxembourg and Sarah Andersen v. Stability AI illustrate the evolving legal landscape.
6.1 Key Findings
I.
Most jurisdictions around the world do not grant copyright to AI-generated content without proper human involvement, as seen in Naruto v. Slater (2018) & Théâtre D'opéra Spatial (2022) ruled by the U.S. Copyright Office. However, some nations like China and Luxemburg have made their early moves to recognize AI-generated works under certain conditions. This shows the difference of approach for acknowledging the AI-generated work. II.
AI developers and user companies are facing lawsuits for using copyrighted materials for training the AI models without authorization, resulting in claims for infringing copyright, as seen in Getty Images v. Stability AI, 2023, Sarah Anderson v. Stability AI
2023 & French Publisher and authors vs Meta, 2025.
III.
Determining the liability in case of infringement is the need of the hour. The legal
ambiguity around the question of who will be held liable in case of AI-generated copyright infringement needs to be resolved through a global effort. IV.
The EU’s AI Act and China’s legal ruling already show different approaches of regulating AI. Since the existing jurisprudence of copyright law covers only the works created by humans, the lack of uniform global standards causes uncertainties for AI developers, content creators, and business enterprises.
6.2 Future Consideration
As AI continues to evolve, legal frameworks need to be adopted to address these complexities by balancing AI innovation and the rights of the copyright holders and AI innovations. I.
To regulate the AI and determine the liability Global legislative and regulatory efforts, along with the judicial pronouncement, are necessary. II.
To seal the gaps left by the current legal frameworks and also to mitigate the issues of Copyright infringement through AI, a new form of copyright law needs to be adopted. III.
Establishing compulsory licensing agreements between AI companies and copyright holders could provide a structured approach for AI companies to use copyrighted materials in training the AI models while compensating the creators. IV.
Legislative initiatives should be taken for the reason of Transparency and Accountability so that the AI developers disclose the copyrighted material that is used in training the models, aiming to protect the rights of copyright holders and ensure ethical use of the content.
6.3 Final Thought
Thus, it can be concluded that the rapid evolution of AI has outpaced existing copyright law. This technology is here to stay, and its impact on copyright law is unavoidable. As AI is reshaping creativity and innovation, the legal framework must evolve to balance technological progress and IPR protection. The decision made today will set and shape how AI-generated content is going to be treated in the future.