Stealing the Spotlight: the Classic tug-of-war of Innovation with Ownership

1. Introduction

In the realm of technological advancement, the advent of Artificial Intelligence (AI) has brought about a paradigm shift in which previously unthinkable achievements are suddenly becoming possible. ChatGPT and other platforms effortlessly produce contextually relevant responses, and AI systems reproduce famous voices, like Mohammad Rafi's, to deliver fresh versions of Bollywood classics. These innovative features, which are driven by advanced Generative AI models, are completely changing the way humans engage with technology, digest information, and pursue creative expression. However, as AI systems use more and more pre-existing content to produce new results, they lead to important discussions about the copyright infringements therein and their intersection with the artificial world.

One such instance is the Naruto v. Slater. The case also referred to as the "Monkey Selfie," in which the US court decided that animals, including the macaque monkey that took a picture, are not entitled to copyrights. This is an example that shows the challenge of crediting non-human entities (like a computer or an animal) as a "creator" or "author". In India, an AI system called "Raghav", autographed artwork "Suryast" produced by him, filed the case to the Copyright Office, seeking copyright for it. So, for starters, after an initial determination that an author had done the work, but the work wasn't human enough to qualify for some protection, a co-author human was found and added to the work, and the work was protected. A few days later, a letter of withdrawal was sent from the copyright office. This ever-present ambiguity in domestic and international legal systems underscores the need to closely examine how intellectual property laws and AI technologies interact. When it comes to supporting new technology, there is a need to combine protection for the rights of creators with encouragement for technological innovation, to ensure that innovation is not made at the expense of creative ownership.

2. Understanding The Stakeholders And Challenges

Now with AI becoming part and parcel of multiple industries, it must identify the key stakeholders:

• Generative AI Developers: This includes those who have companies or people who are creating AI systems that can generate content from potentially some part of copyrighted material.

• Copyright Holders: These are creators or organizations who own the rights to original works that may be used as training data for AI models.

• End Users or Consumers: If the AI-generated content one consumes comes from unprotected sources, users could unintentionally violate copyrights.

The intersection of AI technology and copyright law has created tensions, with major concerns including unauthorized use of data, ethical issues in content creation, and the fear of stifling innovation due to overly strict copyright protections.

3. Key Issues At The Intersection Of AI and Copyright

   • Copyright as an Obstacle to Innovation: “The more artistic protection is favored; the more technological innovation may be discouraged” Stricter copyright protections can sometimes hinder technological advancements. Large Language Models (LLMs) rely on extensive training datasets, often including copyrighted works. Restrictive copyright laws could limit the development of these technologies.

   • Unauthorized Data Harvesting: Unauthorised scraping of copyrighted material from websites or databases by artificial intelligence engineers raises ethical and legal questions. Additionally, "AI data laundering" adds another layer of complexity to transparency because it hides acquired data into something that appears real.

   • Costly Legal Battles and Moral Rights: Copyright holders might be hesitant to go to court because it costs a lot and takes a long time.  Moreover, AI-generated content can infringe on the moral rights of creators, such as their right to be credited and to protect the integrity of their work. It includes the Right to Paternity and Right to Integrity , among others.

   • Liability for End Users: If a consumer uses AI-generated content for commercial reasons without knowing where it came from, they may be breaking the law.
     

     4. International Approaches To AI and Copyright

As the relationship between AI and copyright law is complex at all levels of regulation, it is inherently multi- or even transnational. With the rapid development of AI technologies within the intellectual property (IP) domain, AI training and undeveloped infringement are inescapable issues most countries are struggling to address through their law-making faculties. Since much AI is developed across borders, there have been calls for a consolidated global approach to creating solutions for copyright infringement. International organizations such as the World Intellectual Property Organization (WIPO), for example, are holding talks about how to create a unified legal system that prevents the use of copyright topics without permission while advancing technology.

Different countries have implemented different legal mechanisms to achieve it. As for the United States, safe harbor provisions exist from the Digital Millennium Copyright Act (DMCA) that protect platforms that comply with takedown requests. On the other hand, in Europe, among others, the EU has adopted a more severe stance for its copyright directions with tighter controls on sui generis database rights. On the other hand, in places like the UK and Canada, the legal framework contemplates the notion of "fair dealing," allowing limited usage of copyrighted content in certain situations.

Still, there is no binding global policy regarding AI and how it deals with copyright. Such a split causes an inconsistency, making life harder for AI to comply with different requirements in different regions, and then train the AI models from various datasets. Without a single global standard, it creates legal grey areas and puts AI developers and end-users at risk of expensive litigation. We will only overcome these challenges together by creating the framework within which we will be able to preserve and protect the innovation potential while still protecting our intellectual property on an international level. This would strike a balance that prevents the harmful misuse of copyrighted content all the while allowing AI technologies to develop responsibly around the world.

5. Proposed Policy Solutions

To alleviate these rising concerns, here are a few policies that can help maintain a balance between innovation and writers' and creators' rights:

• Hybrid licensing approach:

A licensing model that combines both ex-ante (before the fact) and ex-post (after the fact) mechanisms can be useful in governing the use of copyrighted material. Pre-licensing would define terms on the front end before usage while post-licensing could permit modifications as necessary.

• Collective management platform:

Establishing a centralized portal between copyright managers and AI developers i.e. IP-CAD could largely simplify the licensing process. Collaboration of stakeholders on an integrated platform generates efficiencies in gaining permission and finalizing terms. CMOs can also extend their efforts by being good at licensing and educating stakeholders.

• Subscription-Based Tiered Pricing:

A tiered payment system could allow artificial intelligence makers access to intellectual resources based on their degree of use. This approach would give suitable compensation for copyright holders and allow flexible access to materials, much like subscription models used by companies like Microsoft 365.

• Supporting AI Startups with Incentives:

44% of private sector companies plan to invest in AI systems in 2023, thereby showing tremendous industry development. Though they sometimes have limited budgets, startups are essential for fostering artificial intelligence creativity. Providing fiscal incentives to generative artificial intelligence companies would inspire them to follow licensing guidelines and lower the possibility of upcoming legal conflicts. This aligns with initiatives like Startup India, which supports research and development through tax credits.

• Enforcing anti-trust regulations:

Big artificial intelligence firms like Google and OpenAI have been charged with monopolizing data. Implementing anti-trust regulations can prevent data monopolies and ensure a fair playing field for smaller AI developers.

• Providing consumer protection:

An indemnification provision can help protect consumers who unintentionally use AI-generated content violating copyright. Like Microsoft's goal of shielding its consumers against copyright litigation, this would guard end users from legal concerns. “If you are challenged on copyright grounds, we will assume responsibility for the potential legal risks involved”

• Opt-Out and Content Removal Rights:

Intellectual property owners should be able to decide how AI systems use their work. ‘Humans Forget, but machines don’t’.  A "Right to Be Removed" clause would let authors ask specifically for the removal of their material from AI training sets in particular circumstances, such as privacy issues. Periodic renewal of licensing agreements may also include an opt-out clause opt-out provision, allowing copyright holders to revoke permission should necessary.

Conclusion

Generative AI is transforming industries and reshaping the creativity landscape, so balancing proprietary rights through a long-term framework within intellectual property rights with the cultivation of generative technology innovation is of critical importance. AI and copyright law are a tricky pair, where too much protection would virtually kill progress, but too little would virtually kill creation. By incorporating policies in a very well thought-through way, from a hybrid licensing model to a collective management platform, as consumer protection and anti-trust regulations, we can implement a system that favors innovation whilst rolling out the carpet for original content creators that have their contribution compensated and recognized. In the end, however, a middle ground will keep AI and human creativity living in peaceful coexistence, advancing with progress, but still protected as intellectual property because after all, and I think Steve Jobs said it best,

“Innovation is the ability to see change as an opportunity, not a threat.”