The advent of generative artificial intelligence (AI) has revolutionized creativity by making it possible for anybody to quickly produce articles, music compositions, or visuals with little to no guidance. Businesses are using ChatGPT, DALL-E 2, and Jasper Art as well as other platforms to leverage Al for their creative requirements. But this change brings with it the legal dangers related to information created by Al. Professionals need to consider these hazards carefully, including attorneys, marketers, and artists.
Although generative AI platforms enable users to produce material quickly, they also bring with them additional legal complications. This paper looks into important legal viewpoints and issues surrounding content created by Al.
Al-generated content frequently uses data that already exists. This paper studies the order to prevent violating the copyrights, trademarks, publicity, and privacy rights of third parties, proper credit is necessary. Users are required to adhere to credit guidelines and comprehend the source of content provided by Al. It will also study the potential copyright and trademark issues must be addressed in the context of AI-generated material. In order to avoid violating the rights of other parties, this study highlights how important it is to make sure AI- generated products comply with legal requirements. Specifically, permits or licenses should be checked when incorporating parts from pre-existing works.
This paper will also study that how can people properly manage the liability of AI-generated material, taking into account the significance of professional evaluation of possible risks[1], careful scrutiny of platform rules, and legal advice? What’s the role of pre-existing laws and does they comply with AI generated content or not? Furthermore, how can the legal issues brought about by AI’s impact on the creative industry be fully comprehended and lessened?
KEYWORDS
AI-Generated Content, Liability, Legal Perspective, Privacy and AI, Copyright Challenges, Ownership of AI- Generated Work.
INTRODUCTION
[1] Christiane Wendehorst, Liablity for Artificial Intelligence, Cambridge University press, (Mar. 11, 2024, 8.45 P.M.), https://www.cambridge.org/core/books/cambridge-handbook-of-responsible-artificial-intelligence/liability-for-artificial- intelligence/12A89C1852919C7DBE9CE982B4DE54B7
What is artificial intelligence (AI) is the first thing that comes to mind when we hear about it. The term artificial intelligence (AI) describes the creation of computer systems that are able to carry out operations that normally require human intelligence, such as decision-making, problem-solving, and natural language comprehension. As far as we are aware, AI has sped up and eased every task but there is a dark side to its use, though. The question that has to be answered is: who is more at fault—AI or the people who use it? AI-generated content raises challenges in determining ownership and copyright[2]. It may be unclear whether the AI creator, the user, or the developer holds rights, leading to legal complexities. The lack of clear authorship in AI-generated content can result in ambiguous attribution, potentially triggering legal disputes over proper credit and intellectual property. AI systems may unintentionally produce content that contains sensitive or private information, which could result in ethical quandaries and privacy violations. Because AI can create material so easily, there are worries about how it might be abused to produce malevolent, false, or deepfake content. Compared to AI-generated content, human-created content usually has clear authorship and well-defined copyright, easing legal aspects. Generative AI systems like ChatGPT, helps everyone through their creative work but there is a cost like what if AI generates similar work as it works on existing content? It may happen that users may lose exclusive rights if AI produces something that is identical to already published works. A fair and efficient system must strike a balance between promoting innovation and defending users’ rights.
Andersen v. Stability AI Ltd., 23-cv-00201-WHO (N.D. Cal. Oct. 30, 2023)[3]
Three artists filed lawsuits against Stability AI, DeviantArt, and Midjourney: Sarah Anderson, Kelly McKernan, and Karla Ortiz. They claimed that these businesses stole their artwork without authorization by using a programme known as Stable Diffusion. It was said that the programme trained computers to produce comparable visuals by using their artwork. The artists said that this amounted to stealing their creations.
After considering the artists’ claims of copyright infringement, the court found that the corporations had improperly utilized the artists’ works. The court decided that the claims for one business, Stability AI, could go forward. Regarding DeviantArt and Midjourney, however, the court ordered the artists to provide more details and allowed them an opportunity to address their concerns.
This case has ignited a heated debate over the intersection of artificial intelligence (AI) and copyright infringement.[4]
RESEARCH METHODOLOGY
The study approach for “Navigating Liability in AI-Generated Content: Legal Perspectives and Challenges” includes a thorough examination of secondary sources in order to clarify the current state of the law pertaining
[2] Gil Appel, Juliana Neelbauer, and David A. Schweidel, Generative AI has an Intellectual Property Problem, HBR, (Mar. 11, 2024, 9.30P.M.) https://hbr.org/2023/04/generative-ai-has-an-intellectual-property-problem
[3] Andersen v. Stability AI Ltd., 23-cv-00201-WHO
[4] Aamir Sheikh, How do Aritists Tackle the Confluence of AI and Copyright, Hall Estill, (Mar. 12, 2024, 7.00 A.M.) https://www.hallestill.com/news-insights/copyright-attorney-randy-mccarthy-andersen-v-stability-ai-case
to this content. In this paper, we will examine previously published materials such as articles and court records. Furthermore, a thorough examination of legal databases will be conducted to ascertain pertinent court cases, legislative frameworks, and other published works referring to artificial intelligence and the law, as well as regulatory changes concerning AI-generated content. Our goal in doing this is to comprehend the legal issues and viewpoints around content generated by artificial intelligence.
REVIEW OF LITERATURE
In the legal challenges of Generative AI: Navigating lawsuits and intellectual property rights, published by the law suits[5], briefly explained that neural network-driven generative AI is revolutionizing sectors such as technology, fashion, and retail, but it’s also giving rise to legal disputes. These cases, which involve unfair competition and copyright violations, bring to light issues with the ownership of work created by artificial intelligence. Litigation is affecting major participants, including startups and tech giants, which is affecting the use of AI. Even if judges have been thoughtful, decisions such as fair use in AI model training influence precedents. These examples highlight the necessity of seeking expert legal advice and the need to review copyright regulations. Analyzing case studies involving Microsoft, OpenAI, and other companies provides useful insights into legal issues and their resolution, helping businesses navigate this complicated terrain.
Makam Ganesh Kumar and Ritam Dutta in their research paper on AI- generated Creations: Navigating and Crafting Effective Policy Frameworks[6] offers a thorough examination of the legal ramifications for AI- generated works of intellectual property (IP). It examines how AI affects copyright, patent, and trademark laws, addressing concerns about authorship, ownership, and content protection for content created by AI. The assessment draws attention to the difficulties in establishing whether AI-generated works are copyrightable and suggests possible remedies, such as acknowledging AI as a legitimate author and modifying current legal frameworks to accept AI-generated intellectual property. It also covers new topics including AI-generated art, AI-assisted inventions, and trademark rights enforcement in relation to AI-generated trademarks.
Gil Appel, Juliana Neelbauer, and David A. Schweidel in Generative AI has an Intellectual Problem[7] explained that content creation is being revolutionised by Generative Artificial Intelligence (AI), as evidenced by platforms such as DALL·E 2 and Stable Diffusion that produce amazing text and imagery. But there are legal issues with this innovation, especially with relation to intellectual property (IP) rights.
The intersection of generative AI with current legal frameworks gives rise to discussions about ownership, usage rights, and infringement of AI works. The fair use doctrine is crucial because it permits the
[5] Mellisa H. grey, The Legal Challenges of Generative AI: Navigating Lawsuits and Intellectual Property Rights (Mar. 12, 2024, 9.29A.M.) https://thelawspot.com/the-legal-challenges-of-generative-ai-navigating-lawsuits-and-intellectual-property-rights/
[6] Maken Ganesh Kumar and Ritam Dutta, AI-generated Creations: Navigating Legal Implications and Crafting Effective Policy Frameworks, SSRN (Aug. 3, 2023), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4520938#:~:text=These%20AI%2Dgenerated%20works%20challenge,the%2 0realm%20of%20intellectual%20property
[7] Gil Appel, Juliana Neelbauer, and David A. Schweidel, Generative AI has and Intellectual Property Problems, HBR, (Mar. 12,2024, 12.50 P.M.) https://hbr.org/2023/04/generative-ai-has-an-intellectual-property-problem
transformative use of copyrighted material while casting doubt on the validity of derivative works. A precedent was set by the Google search engine scraping case, but the Warhol Foundation case raises more issues about the definition of “Transformative.” This emphasizes the difficulty of information generated by AI.
METHODS
The method employed in this paper involves conducting a comprehensive analysis of previous research papers, academic publications, journals, blogs, and other relevant literature. Additionally, the liability associated with AI-generated content is addressed, as well as the issues raised by it, using precedents, relevant statutes, Supreme Court decisions, and other legal framework.
EXPLORING LEGAL IMPLICATIONS OF AI GENERATED CONTENT IN INDIA
While addressing liabilities connected with AI-generated content, it is important to examine different relevant sections and legal rules, particularly in the context of intellectual property law in India, however, there is no specific laws regulating AI in India.[8] But the Ministry of Electronics and Information Technology(MeiTY) is the regulating body of AI in India. It has responsibility development, implementation and management of AI laws in India.9
The National Strategy for Artificial Intelligence
It was released in 2018 by NITI Aayog. It argues for a market-based approach to maximize the potential of artificial intelligence and suggests a two-tiered structure (CORE and ICTAI) to satisfy research goals. Emphasis is placed on security, privacy, and ethical issues; the creation of ethics councils and data protection frameworks is advised. The report emphasizes how crucial worker skill development, government support, and public-private cooperation are to achieving the #AIForAll objective.
Personal Data Protection Bill
The purpose of the Personal Data Protection Bill, 2019 was to safeguard people’s privacy and control the collection, use, and exchange of personal data in India. It suggested creating an Indian Data Protection Authority. It was criticized, nevertheless, for a number of reasons, including gaps in the identification of governmental entities that had access to residents’ data and worries about government access to private information. August 3, 2022 saw the withdrawal of the measure from Parliament.
[8] Rahul Kapoor Partner, Shokoh H. Yahoubi Associate and Theresa T. Kalathil Associate, AI Regulation in India: Current State and Future perspective, Morgan Lewis, (Mar. 12, 2024, 3.40P.M.) https://www.morganlewis.com/blogs/sourcingatmorganlewis/2024/01/ai-regulation-in-india-current-state-and-future- perspectives#:~:text=In%20the%20last%20several%20years,laws%20regulating%20AI%20in%20India
[9] Shreya Banerjee, Laws Related to AI in India, INBA viewpoint, (Mar. 12, 2024, 9.10 P.M.) https://inbaviewpoint.org/laws- related-to-artificial-intelligence- inindia/#:~:text=The%20Ministry%20of%20Electronics%20and,laws%20and%20guidelines%20in%20India
The goal of the Data Protection Bill 2021 was to create stringent breach reporting requirements while extending protection to both personal and non-personal data. But because of flaws, it was withdrawn in August 2022. The Digital Personal Data Protection Bill 2022 was launched by the Indian government in November 2022. It is scheduled to be presented in Parliament during the monsoon session and is available for public comment until January 2023.
Digital Personal Data Protection Act, 2023
Building a strong data privacy framework, highlighting the role of significant data fiduciaries (SDF), and endowing citizens with particular data principle rights are the main objectives of India’s Digital Personal Data Protection (DPDP) Act, 2023. The severe consequences of non-compliance, which can reach INR 250 crore, emphasize how important compliance is. Since the legislation excludes some data categories and is projected to have a substantial impact on specific sectors, it is imperative that robust data privacy implementation strategies be developed.
Information Technology Act, 2000
Although IT Act does not explicitly mention AI, specific provisions within the Act are applicable to AI-related activities. Section 43A of the IT Act enables compensation in case of a breach of data privacy resulting from negligent handling of sensitive personal information. This provision is particularly relevant in the context of AI systems that process user data. Another provision is Section 73A of this act[10]. Information Technology (Intermediary Guidelines and Digital Media Ethics Codes) Rules 2021 obligates the social media platforms to exercise greater diligence regarding content on their platforms.[11]
The Copyright Act, 1957
the copyright act of 1957 safeguards original works of authorship against unauthorized usage; however, it does not extend its definition of an author to include AI generated content.
The Patent Act, 1970
Under the patent act, protections are afforded to new inventions and discoveries that are new and non-obvious; nevertheless, AI generated creations fall outside the scope of its coverage.
The Trade Marks Act, 1999
Regarding trademarks, AI’s absence of legal personality precludes it from being recognized as an owner of trademarks.
CASE LAWS
[10] Diya Saraswat, Laws Governing AI in India: Everything you Should Know, Legal Service India E-Journal, (Mar. 12, 10.09 P.M.) https://www.legalserviceindia.com/legal/article-13111-laws-governing-ai-in-india-everything-you-should-know.html
[11] Shreya Banerjee, Laws Related to AI in India, INBA viewpoint, (Mar. 12, 2024, 11.50 P.M.) https://inbaviewpoint.org/laws- related-to-artificial-intelligence-in- india/#:~:text=The%20Ministry%20of%20Electronics%20and,laws%20and%20guidelines%20in%20India
Now, let’s gain clearer insights through the analysis of some relevant cases –
Eastern Book Company v. D. B. Modak
Eastern Book Company and EBC Publishing Private Limited, engaged in legal book printing, published ‘Supreme Court Cases’ (SCC) since 1969, incorporating formatting, numbering, and headnotes to make it user- friendly. Spectrum Business Support Limited and Regent Datatech Private Limited released ‘Grand Jurix’ and ‘The Laws,’ allegedly copying SCC onto CD-ROMs, infringing on the appellants’ IP rights.
The Supreme Court ruled that derivative works have to show more than just a simple copy of the original in order to be granted copyright protection. The court granted the Petitioners exclusive rights to their content after determining that their work exhibited sufficient originality, based on the Canadian “skill and judgement” criteria. The Petitioners’ copyright claims were upheld when the Respondents were ordered to stop replicating particular components.
Bishwanath Prasad Radhey Shyam v. Hindustan Metal Industries
In 1951, the respondent invented a novel technique for producing cutlery that increased productivity and security. A lawsuit resulted in 1952 when the appellant was discovered to be exploiting the patented process. The appellant contended that the patented method lacked originality and an inventive step.
The trial judge dismissed the lawsuit and permitted the revocation of the patent after finding that the invention lacked novelty and inventive step. Noting the uniqueness and usefulness of the invention, the Appellate Bench disagreed. Overturned the trial court’s ruling and maintained the patent’s legality. Stressed the significance of utility and creative step in patent law.
Amar Nath Sehgal v. Union of India
In the 1960s, Delhi’s Vigyan Bhawan had a mural sculpture made by sculptor Amar Nath Sehgal. After being purchased by the Indian government in 1979, the sculpture was stored, which led to damage. Accusing the government of violating his moral rights as an artist, Sehgal filed a lawsuit.
Sehgal’s moral rights over the artwork were recognized by the Delhi High Court, which ruled in his favor. The court decided that even after his work was purchased by the government, Sehgal was entitled to maintain proprietary rights over it. It concluded that Sehgal’s moral rights had been violated by the government’s activities, which included causing harm to the artwork. A payment of Rs 5 lakh was given to Sehgal, and portions of the artwork were sent to him for conservation and possible commerce.
Moral rights are related to the feeling and emotions of a human author. These rights are not meant for an AI.[15]
[12] Eastern book company & Ors. v. D.B. Modak & Anr., AIR 2008 SC 809
[13] Biswanath Prasad Radhey Shyam v. Hindustan Metal Industries, AIR 1982 SC 1444
[14] Amar Nath Sehgal v. Union of India, 117(2005)DLT717
[15] V.K. Ahuja, Artificial Intelligence and Copyright: Issues and Challenges, ILI Law Review, 277, (2020) https://ili.ac.in/pdf/vka.pdf
M/S Kibow Biotech v. M/S The Registrar of Trademarks
A decision was made in this case that an AI system cannot be regarded as the owner of a trademark, according to the Registrar of Trade Marks. The court determined that only a person is eligible to apply for and be registered as the owner of a trademark under the Trade Marks Act, 1999.
Gaurav Bhatia v. Union of India
In this case the Delhi High Court ruled that AI-Generated inventions could be patent if they meet the requirements of novality, non-obviousness, and industrial applicability under the patent act.
CHALLENGES
Apart from the difficulties while figuring out who truly owns the art, music, and various other AI generated content; additional challenges for AI generated content includes-
- Deepfakes
Deepfakes resemble computer-generated magic. Artificial Intelligence (AI) is used to create synthetic images, films, and even voices that mimic the appearance and sound of the original. Why is this such a big deal now? Let’s say someone fabricates a video purporting to show a famous person saying something horrible or false. Because it appears so genuine, people may accept it and their opinions of that person may shift. That cannot be fair, can it? It resembles a cruel practical joke, although considerably larger in scope. Accordingly, deepfakes can lead to numerous issues. They can disseminate false information, perplex people, and even tamper with crucial events. People are concerned about them and attempting to determine ways to prevent them from causing problems because of this.
Anil Kapoor v. Simply Life India & Ors
Anil Kapoor, the plaintiff, claimed a number of infringements, including unfair competition, copyright for lines like “jhakaas,” and personality rights. The defendants sold products featuring Kapoor’s image, produced deepfakes, and registered domain names using his identity.
The court found in Kapoor’s favour, highlighting the need to defend celebrity rights from unauthorised commercial usage and identity theft. It was declared unlawful to use Kapoor’s name or likeness on any kind of product or internet. The Copyright Act and Trademark Act’s protection was emphasised,
[16] Vishakha Aditya, Navigating the Legal Challenges Posted by AI on Intellectual Property, Legal Service India E-Journal, (Mar. 13, 09.08 A.M.), https://www.legalserviceindia.com/legal/article-12029-navigating-the-legal-challenges-posed-by-ai-on- intellectual-property.html
[17] Vishakha Aditya, Navigating the Legal Challenges Posted by AI on Intellectual Property, Legal Service India E-Journal, (Mar. 13, 10.15 A.M.), https://www.legalserviceindia.com/legal/article-12029-navigating-the-legal-challenges-posed-by-ai-on- intellectual-property.html
[18] Anil Kapoor v. Simply Life India & Ors CS 9(COMM) 652/2023
enabling celebrities to trademark their appearances and speech. While defences against personality rights claims were taken into consideration, they were deemed insufficient in this instance.
- Defamation because of AI-Generated Content
Artificial intelligence (AI) systems have the ability to produce content, such as articles or social media posts, that may unintentionally contain inaccurate information or defamatory statements. This is a problem since, typically, human authors are involved in defamation. It becomes difficult to assign blame for malicious content produced by AI. Who is responsible for the data entry—the user, the AI, or the AI creator?
- Data Privacy
People provide a great deal of personal information to AI systems. This might reveal your location, your browsing history, or even just your name. Concerns exist around the use and security of the vast amount of data being gathered.
Justice K.S. Puttaswamy (Retd.) & Anr v. Union of India
in this case, the Supreme Court of India, overruled M.P. Sharma and Kharak Singh case and declared that privacy is a fundamental right under Article 21 of the Constitution.
- Consent and Profiling
AI has the potential to violate your private rights when it makes judgments or creates profiles of individuals (e.g., predicting what you might like to buy based on previous purchases). You might not even be aware that AI is making judgments on your behalf at times. This may not be true or fair, and you may not have agreed to it, so these are worrying possibilities. Encouraging individuals to know what is happening with their data and to have a voice in its usage is therefore a challenge.
SUGGESTIONS
To manage situations involving AI-generated content, legal systems must change. They need to think about who should be held responsible and how to stop or deal with AI-caused defamation. There is an urgent need to reform the current regulations since they are ill-suited to address the myriad of issues that artificial intelligence brings about. To properly handle the complexities of AI-generated content and manage liability more effectively, the existing laws, such as the Trademark Act, Patent Act, and Copyright Act, need to be revised.
Thus, it’s imperative to have clear policies on accountability and attribution. These rules ought to outline who is in charge of AI-generated material and how accountability is to be assessed.
It is also crucial that you read and comprehend the terms of the website or tools before using the AI for generating anything. Usually, the terms and conditions or terms of services contain these guidelines.
[19] Justice K.S. Puttaswamy (Retd.) & Anr. V. Union of India & Ors., AIR 2018 SC (SUPP)1841
If AI-generating tools are using pre-existing information then the proper credit must be attributed to the individual or entity responsible for the original idea or content.
CONCLUSION
In conclusion, there are many obstacles in addressing concerns like bias, responsibility, and intellectual property rights when it comes to Artificial Intelligence (AI) because there aren’t enough comprehensive laws, clear ethical standards, or appropriate regulations. Despite these drawbacks, India is working to create an efficient framework for the regulation of AI, stressing the necessity of changing laws often to keep up with the rapid evolution of technology. Governments and legislators must collaborate in order to close these gaps and guarantee that AI is applied morally and sensibly for the good of society.
Ankita Maurya
Graphic Era Hill University, Dehradun