Artificial intelligence (AI) has emerged as both a transformative technology and a contentious legal frontier, particularly in the realm of intellectual property (IP). Recently, AI developers have secured significant victories in copyright litigation, notably with OpenAI's triumph against claims brought by Raw Story and AlterNet. These developments underscore the complexities of AI's relationship with copyright law and its broader implications for creators, corporations, and the legal landscape. The ongoing legal disputes and case laws are poised to shape the future of copyright in the AI era.
OpenAI's Legal Victory: A Landmark in AI Copyright Jurisprudence
In November 2024, a U.S. District Court dismissed claims by Raw Story and AlterNet against OpenAI under the Digital Millennium Copyright Act (DMCA). The plaintiffs alleged that OpenAI violated DMCA Section 1202(b) by stripping copyright management information (CMI) from their articles while training its large language models (LLMs). They argued that this facilitated copyright infringement through ChatGPT's outputs. However, Judge Colleen McMahon ruled that the plaintiffs failed to demonstrate a "concrete injury-in-fact," a critical threshold under Article III of the U.S. Constitution.
Central to the court's decision was the absence of tangible harm. While the plaintiffs claimed that their works were used to train ChatGPT, they could not prove that the AI model reproduced or disseminated their content verbatim. This ruling aligns with the U.S. Supreme Court's 2021 decision in TransUnion v. Ramirez, which emphasized that plaintiffs must show concrete harm rather than relying solely on statutory violations. This precedent imposes a significant burden on plaintiffs in AI-related copyright cases, potentially reshaping the litigation landscape.
Implications for Copyright Owners and AI Developers
The dismissal of the Raw Story case highlights the challenges copyright owners face in holding AI developers accountable. Demonstrating direct infringement by AI models requires concrete evidence, such as outputs closely resembling copyrighted works. However, generative AI often synthesizes information rather than reproducing content verbatim, complicating claims of infringement.
For AI developers, the ruling reinforces the importance of fair use and related doctrines as legal defences. OpenAI's argument that its models are trained on publicly available data, supported by long-standing legal precedents, was pivotal in securing this victory. Yet, this reliance on fair use remains contentious, with critics arguing that it undermines creators' rights and the economic value of their works.
Case Law: A Global Perspective
The ongoing legal disputes surrounding generative AI and copyright are highlighting the diverse approaches taken by different jurisdictions in regulating the use of copyrighted materials in AI training. In India, a landmark lawsuit filed by Asian News International (ANI) against OpenAI has brought attention to the unauthorized use of news content for training ChatGPT. ANI alleges that OpenAI not only used its copyrighted material without consent but also generated outputs that reproduced its work verbatim, potentially harming its reputation and spreading misinformation. This case, currently before the Delhi High Court, could set significant precedents for AI-related copyright issues in India. Similarly, in the United States, Getty Images has filed a lawsuit against Stability AI for using millions of watermarked images from its database without permission in the development of its Stable Diffusion model. Stability AI defends its actions under the fair use doctrine, raising key questions about the legal and ethical boundaries of AI data acquisition.
Other notable cases, such as Andersen v. Stability AI and Tremblay v. OpenAI, further illustrate the challenges surrounding AI’s use of copyrighted content. In Andersen, three artists sued Stability AI for allegedly replicating their artistic styles without permission, with some claims surviving pretrial motions, signaling courts' willingness to explore AI and derivative works laws. In Tremblay, plaintiffs claimed OpenAI removed copyright management information from their works, but the court dismissed the case for lack of evidence. The Kadrey v. Meta Platforms case similarly dealt with whether AI models constitute derivative works of copyrighted data used for training, but the court ruled that such claims were "nonsensical." These cases collectively highlight the difficulties plaintiffs face in proving infringement or harm and underscore the broader challenge of adapting copyright law to the complexities of AI technologies.
The Fair Use Doctrine: A Contested Shield
At the heart of these disputes lies the fair use doctrine, which allows limited use of copyrighted material for purposes such as criticism, commentary, and research. In the context of AI, courts must grapple with whether training models on copyrighted content constitutes transformative use. Critics argue that AI's reliance on vast datasets risks commodifying creative works without adequate compensation to their creators.
A parallel can be drawn with the U.S. Supreme Court's decision in Google v. Oracle, where Google's use of Java APIs was deemed fair use due to its transformative nature. However, the application of this principle to generative AI remains unsettled, with courts divided on whether the outputs of such models sufficiently transform the original works.
Ethical Considerations and Future Challenges
Beyond legal doctrines, the ethical implications of AI's use of copyrighted material cannot be ignored. Creators argue that the unlicensed use of their works for training AI systems devalues their labor and threatens their livelihoods. Meanwhile, AI developers contend that access to diverse datasets is essential for innovation and the democratization of technology.
This tension raises broader questions about the balance between protecting IP rights and fostering technological advancement. Should AI developers be required to obtain explicit consent and compensate creators for using their works? Or does the principle of fair use provide sufficient safeguards against exploitation?
The Way Forward: Bridging the Gap
To address these challenges, a multi-faceted approach is needed. Policymakers must clarify the legal framework governing AI and copyright, ensuring consistency across jurisdictions. This could involve establishing guidelines for permissible data usage and fair compensation mechanisms for creators. AI developers should prioritize transparency in their training processes, including providing metadata on the sources used. This would allow copyright owners to verify whether their works have been included and seek redress if necessary. Additionally, courts must adapt to the complexities of AI-related cases by adopting a nuanced approach to fair use and concrete harm, considering the broader societal benefits of AI while safeguarding the rights of creators.
The legal battle between AI developers and copyright owners is far from over. While OpenAI's recent victory marks a milestone, it also underscores the evolving nature of copyright law in the AI age. As courts, policymakers, and stakeholders navigate this uncharted territory, the need for a balanced and equitable framework becomes increasingly urgent. Only by addressing these challenges can we ensure that AI serves as a tool for innovation without undermining the rights and livelihoods of creators.
IP Round up
Copyright Clash: Nayanthara vs. Dhanush Over Netflix Clip: A legal battle has erupted between actors Nayanthara and Dhanush over a three-second clip from the 2015 film Naanum Rowdy Dhaan featured in Netflix’s documentary Nayanthara: Beyond the Fairytale. Dhanush, the film’s producer, claims copyright infringement and has issued a legal notice demanding the clip’s removal within 24 hours, threatening ₹10 crore in damages against Netflix, Nayanthara, and her husband, filmmaker Vignesh Shivan. While Nayanthara criticized Dhanush’s actions as excessive in an open letter, his legal team argued the clip belongs to the production team and challenged her to provide evidence to the contrary. (Source: Deccan Chronicle)
Britannia Secures Victory in Good Day Trademark Battle:
The Delhi High Court ruled in favor of Britannia Industries in a trademark infringement case against Desi Bites Snack, protecting its iconic GOOD DAY trademark. Desi Bites had used the mark for products like soan papdi and papad without authorization. The court issued an ex parte injunction, prohibiting the use of the GOOD DAY trademark and directing the removal of infringing products from all platforms. Britannia argued that its trademark, registered since 1986, represents significant goodwill in the bakery and confectionery sector, with annual sales of over ₹16,000 crore and a vast retail network. The court recognized the risk of consumer confusion and irreparable harm, concluding that Desi Bites’ actions constituted trademark infringement, passing off, and unfair competition, and upheld Britannia's claim of brand dilution. (Source: CNBCTV18)
Delhi High Court Summons OpenAI Over Alleged Copyright Infringement: The Delhi High Court recently issued summons to OpenAI, the company behind ChatGPT, in response to a copyright infringement lawsuit filed by news agency ANI. This is the first legal action against OpenAI in India, where ANI claims that ChatGPT misattributed political news to the agency, potentially spreading misinformation and causing public unrest. The court refrained from granting any immediate injunction, recognizing the complexity of the case, and appointed an amicus curiae for assistance. ANI argued that OpenAI had no right to exploit publicly available news content without permission, especially when it led to false attribution. OpenAI, however, denied the allegations, stating that copyright laws protect expressions, not facts, and emphasized that its operations do not involve content storage or reproduction within India. The case is scheduled for further hearing in January 2024. (Mint)
USPTO Unveils New Trademark Fees effective from 2025:
In May 2023, the U.S. Patent and Trademark Office (USPTO) proposed new fees for trademark filings, which were finalized on November 15, 2024. These changes, effective January 18, 2025, aim to streamline filing procedures and improve efficiency by imposing surcharges for incomplete applications. A new base filing fee of $350/class will replace existing TEAS options, with additional charges for incomplete or free-form descriptions of goods and services. Madrid filings will see a fee increase instead of surcharges due to WIPO's limitations. The USPTO has also raised post-registration and petition fees to cover increased costs. These adjustments aim to align fees with the actual costs of trademark services, encouraging more complete and timely filings while adding strategic considerations for applicants. The changes could complicate budgeting and filing strategies, especially for businesses in emerging sectors. (Dykema)
Grow Digi's Blockchain Patents Boost Efficiency and Security: Indian blockchain startup Grow Digi has received two patents for technology that reduces data storage by 95%, improves transaction speeds, and enhances security. The company's Layer 1 blockchain, which serves as the foundation for enterprise applications, decreases transaction times from five seconds to under 300 milliseconds. It also introduces randomized transaction verification to prevent collusion within the blockchain. Founded in 2021, Grow Digi has developed blockchain solutions for industries like electric vehicles and is working on health applications, allowing patients to control their medical data. The company plans to monetize its platform and onboard customers. (Business Standards)
References:
Agrawal, A. (2024, November 19). News agency ANI sues OpenAI for alleged violation of intellectual property rights. Hindustan Times. Retrieved from https://www.hindustantimes.com/india-news/news-agency-ani-sues-openai-for-alleged-violation-of-intellectual-property-rights-101731945177550.html
Ahmad, M. S. (2024, November 18). The beginning of AI copyright litigation in India: ANI vs. OpenAI in DHC tomorrow. SpicyIP. Retrieved from https://spicyip.com/2024/11/the-beginning-of-ai-copyright-litigation-in-india-ani-vs-openai-in-dhc-tomorrow.html
Authors Alliance. (2024, September 3). The AI copyright hype: Legal claims that didn’t hold up. Authors Alliance. Retrieved from https://www.authorsalliance.org/2024/09/03/the-ai-copyright-hype-legal-claims-that-didnt-hold-up/
Brittain, B. (2024, November 8). OpenAI defeats news outlet’s copyright lawsuit over AI training data. Reuters. Retrieved from https://www.reuters.com/legal/litigation/openai-defeats-news-outlets-copyright-lawsuit-over-ai-training-now-2024-11-07/
Dykema. (2024, November 18). Upcoming changes to trademark filing fees: A guide to the U.S. Patent and Trademark Office’s new regulations. Dykema. Retrieved from https://www.dykema.com/news-insights/upcoming-changes-to-trademark-filing-fees-a-guide-to-the-us-patent-and-trademark-offices-new-regulations.html
Ha, A. (2024, October 19). Penguin Random House is adding an AI warning to its books’ copyright pages. TechCrunch. Retrieved from https://techcrunch.com/2024/10/19/penguin-random-house-is-adding-an-ai-warning-to-its-books-copyright-pages/
Knibbs, K. (2024, November 8). OpenAI defeats copyright lawsuit over DMCA standing with AlterNet, Raw Story. Wired. Retrieved from https://www.wired.com/story/opena-alternet-raw-story-copyright-lawsuit-dmca-standing/
Moss, A. (2024, November 10). Raw Story copyright lawsuit standing. Copyright Lately. Retrieved from https://copyrightlately.com/raw-story-copyright-lawsuit-standing/
Pant, A. (2024, November 18). Britannia secures injunction against Desi Bites for Good Day trademark infringement. CNBC TV18. Retrieved from https://www.cnbctv18.com/business/companies/britannia-secures-injunction-against-desi-bites-for-good-day-trademark-infringement-19510936.htm
PTI. (2024, November 17). Grow Digi gets two patents for sector-agnostic blockchain tech layer. Business Standard. Retrieved from https://www.business-standard.com/companies/news/grow-digi-gets-two-patents-for-sector-agnostic-blockchain-tech-layer-124111700310_1.html
Yadav, K. (2024, November 19). OpenAI: Delhi high court summons news agency ANI in copyright infringement plea over fake news. LiveMint. Retrieved from https://www.livemint.com/companies/news/openai-chatgpt-delhi-high-court-summons-news-agency-ani-copyright-infringement-plea-fake-news-wikimedia-foundation-11732000483630.html