AI and Intellectual Property Law: Reconstructing Ownership, Liability and Commercial Exploitation

Written by Harshkumar Pendharkar and Pranjali Paliwal, the authors are a law students currently pursuing BA.LLB from Symbiosis Law School, Pune

“The law must be stable, but it must not stand still” - Roscoe Pound

Introduction

The advent of Artificial Intelligence (AI) has precipitated a paradigmatic shift in the landscape of Intellectual Property Rights (IPR) and Commercial Law, necessitating a comprehensive re-evaluation of existing legal frameworks. The confluence of AI technologies with traditional IPR principles has engendered novel jurisprudential challenges that require urgent legislative attention and judicial interpretation. As AI systems become increasingly sophisticated, capable of generating creative works, inventions, and commercial solutions autonomously, the fundamental premises upon which IPR law is predicated are being subjected to unprecedented scrutiny. The commercial implications of AI-generated intellectual property extend far beyond academic discourse, permeating industries ranging from pharmaceuticals to entertainment, thereby necessitating a robust legal framework that can accommodate the unique characteristics of AI-generated content while preserving the incentive structures that underpin innovation.

The Intersection of AI and Traditional IPR Frameworks

Patent Law and AI Inventions

The traditional patent law framework, predicated on human inventorship, encounters significant challenges when confronted with AI-generated inventions. Statutory inventorship, as codified in patent laws across the world, assumes human intent and creativity. The United Kingdom's case law in Thaler v. Comptroller-General of Patents, Designs and Trade Marks is a classic example of judicial hesitation to validate AI systems as inventors. The Court of Appeal upheld the decision that DABUS, an AI system, could not be considered an inventor under the Patents Act 1977, reasoning that inventorship requires a natural person. This precedent has been influential across jurisdictions, with similar decisions emerging in the United States. and Australia. However, the jurisprudential landscape remains fragmented. The European Patent Office’s position, as articulated in its Guidelines for Examination, maintains that an inventor must be a natural person, thereby excluding AI systems from inventorship. However, there are jurisdictions like South Africa that have shown more liberal inclinations by issuing patents if the inventor is an AI system.

Copyright Law and AI-Generated Works

The application of copyright law to works produced by AI also poses equally challenging questions. The underlying principle of the copyright law to protect original works of authorship presumes human expression and creativity. The emergence of AI systems capable of generating literary, musical, and artistic works autonomously has necessitated a re-examination of the authorship requirement.

The Indian Copyright Act’s definition of “author” in relation to literary, dramatic, musical, and artistic works presupposes human agency. However, the Act’s provision for computer-generated works under Section 2(d)(vi) offers a potential framework for addressing AI-generated content, albeit with limitations regarding the requisite human involvement in the creative process. The landmark case of Narendra Publishing House v. Usha Rani established the principle that copyright subsists in works that demonstrate minimal creativity and originality. This precedent, when juxtaposed with contemporary AI capabilities, raises pertinent questions about the threshold of creativity required for copyright protection and whether AI-generated works can satisfy such requirements.

Trademark Law in the AI Era

The intersection of trademark law and AI is particularly challenging, especially with regard to AI-generated marks and the application of AI to trademark prosecution and enforcement. The established system of trademark law, created to safeguard distinctive signs employed in trade, has to now adapt to situations where AI can create protectable marks. The decision of Booking.com B.V. v. United States Patent and Trademark Office set significant precedents about the distinctiveness of generic terms used in combination with generic top-level domains. This doctrine becomes highly applicable when examining AI-generated trademark filings, where algorithmic means can generate combinations that test the conventional standards of distinctiveness and creativity.

Commercial Law Implications

Contractual Frameworks for AI-Related IP Transactions

The commercial use of AI-generated intellectual property demands advanced contractual structures that account for the novel nature of AI systems and their outputs. Conventional IP licensing contracts, based on human invention and authorship, need significant adaptation to include AI-generated content. The enforceability of AI-generated IP contracts is new in its challenge, especially in relation to the determination of contracting parties and ascribing rights and obligations. The principle in Satyabrata Ghose v. Mugneeram Bangur & Co. concerning the frustration of contracts can be applied in situations where AI systems do not perform as expected, thus impacting the commercial value of AI-generated IP..

Liability and Indemnification Issues

The issue of AI-generated IP infringement liability is a severe commercial law challenge. Conventional doctrines of vicarious liability and contributory infringement, as developed in cases such as Sony Corp. of America v. Universal City Studios, Inc., need to be modified to deal with situations in which AI systems create infringing material on their own. The notion of "reasonable care" in the case of AI systems, as being potentially similar to the standards set in Donoghue v. Stevenson, could serve as a basis for assigning liability in AI IP disputes. However, the autonomous nature of AI decision-making complicates traditional notions of foreseeability and proximate cause.

Valuation and Monetization Challenges

The commercial valuation of AI-generated IP presents unprecedented challenges for traditional valuation methodologies. The income approach, market approach, and cost approach to IP valuation, as established in commercial practice, must be adapted to account for the unique characteristics of AI-generated content.

The case of Roche Products v. Bayer Corporation established important precedents regarding the valuation of pharmaceutical patents, principles that may be relevant when valuing AI-generated pharmaceutical innovations.

Regulatory Responses and International Harmonization

National Legislative Initiatives

Various jurisdictions have initiated legislative responses to address the challenges posed by AI-generated IP. The United Kingdom’s Intellectual Property Office has published consultation documents on AI and IP, while the European Union’s proposed AI Act includes provisions relevant to IP considerations.

India’s National Strategy for Artificial Intelligence acknowledges the need for IP framework adaptation, though specific legislative amendments remain pending.

International Harmonization Efforts

The World Intellectual Property Organization (WIPO) has established a Conversation on AI and IP to facilitate international dialogue on these issues. The organization’s draft issues paper on AI and inventorship provides a comprehensive analysis of the challenges and potential solutions.

The Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) may require amendment or interpretive guidance to address AI-generated inventions effectively.

Commercial Dispute Resolution in the AI Era

Arbitration and Mediation Challenges

The resolution of AI-related IP disputes through alternative dispute resolution mechanisms presents unique challenges. The complexity of AI technologies may require specialized arbitrators with technical expertise, potentially necessitating amendments to institutional arbitration rules.

The case of Bharat Aluminium Co. v. Kaiser Aluminium Technical Service established important principles regarding the seat of arbitration, which may become relevant in cross-border AI IP disputes.

Evidence and Expert Testimony

The presentation of evidence in AI-related IP disputes requires sophisticated understanding of AI technologies and their capabilities. The principles established in State of U.P. v. Krishna Gopal regarding expert testimony may need adaptation to address the unique characteristics of AI-generated evidence.

Future Directions and Recommendations

Legislative Reform Priorities

The urgent need for legislative reform in AI-related IP law cannot be overstated. The primary areas requiring immediate attention encompass the establishment of clear statutory provisions addressing AI-generated inventions and creative works, which would provide much-needed clarity regarding inventorship and authorship frameworks. Sophisticated frameworks for assessing liability in IP disputes involving AI need to be created to handle the singular challenges that autonomous AI systems present. Moreover, standardized methodologies for assessing the value of IP produced by AI are important for their facilitation in commercial transactions and equitable compensation. International coordination through congruence with worldwide initiatives continues to be important to provide congruent standards across jurisdictions and avoid regulatory arbitrage.

Commercial Best Practices

Commercial organizations in the AI sector need to have robust IP strategies in place that include precise agreements on AI-created IP ownership and licensing terms. They need to have the sophisticated due diligence practices that are tailored specifically for AI IP deals, in view of the peculiar technical and legal intricacies. Risk management processes need to have robust insurance and indemnity provisions that consider the new liability risks tied up with AI systems. In addition, continuous compliance monitoring through routine audits of AI systems ensures continuous compliance with IP requirements and also aids in detecting potential infringement issues before these become expensive disputes.

Conclusion

The confluence of AI, IPR, and Commercial Law is one of the greatest challenges to modern legal systems. The conventional paradigms that have regulated intellectual property rights for millennia are being put to their greatest test ever as AI systems illustrate abilities that contest underlying assumptions regarding creativity, inventorship, and authorship.

The business consequences of these challenges reach far beyond the academic literature, impacting industries that are worth trillions of dollars around the world. The call for immediate legislative reform, global harmonization, and advanced commercial practices has never been more pressing. As we embark on this period of change, the insight of Justice Cardozo that “the law never is, but is always about to be” rings especially true. The law needs to adapt to the realities of intellectual property generated by AI while maintaining essential incentive structures that fuel innovation and creativity.

The way forward calls for cooperation among legislators, judiciary, commercial practitioners, and technologists in crafting frameworks that are legally sound and commercially feasible. It is only through such cooperation that we can make sure the law is relevant and effective in the era of artificial intelligence, thus achieving its ultimate goal of facilitating innovation and safeguarding legitimate commercial interests.

The path to thoroughgoing AI-IP legislation is potentially complicated and difficult, but it is necessary to preserving the fine balance between innovation incentives and public interest that has long been the signature of successful intellectual property law. As we come to this intersection of technology and law, choices today will determine the commercial landscape for centuries to come.

References

Books and Articles:

Cardozo, Benjamin N., The Nature of the Judicial Process, 1921

Cases Cited:

Bharat Aluminium Co. v. Kaiser Aluminium Technical Service (2012) 9 SCC 552
Booking.com B.V. v. United States Patent and Trademark Office 591 U.S. ___ (2020)
Commissioner of Patents v. Thaler [2022] FCAFC 62
Donoghue v. Stevenson [1932] AC 562
Narendra Publishing House v. Usha Rani 1964 AIR 1480
Roche Products v. Bayer Corporation 2012 (52) PTC 1 (Del)
Satyabrata Ghose v. Mugneeram Bangur & Co. AIR 1954 SC 44
Sony Corp. of America v. Universal City Studios, Inc. 464 U.S. 417 (1984)
State of U.P. v. Krishna Gopal (1988) 4 SCC 302
Thaler v. Comptroller-General of Patents, Designs and Trade Marks [2021] EWCA Civ 1374
Thaler v. Hirshfeld, 558 F.Supp.3d 238 (E.D. Va. 2021)

Official Documents and Reports:

EPO Guidelines for Examination, Part A, Chapter IV, 4.1
European Patent Office Guidelines for Examination, Part G, Chapter VII, 2024
European Union Proposal for a Regulation on AI, COM/2021/206 final
NITI Aayog, National Strategy for Artificial Intelligence, 2018
South African Patent No. 2021/03242
UKIPO Consultation on AI and IP, 2021
WIPO Conversation on AI and IP Policy, 2019-2022
WIPO Issues Paper on AI and Inventorship, 2020
World Intellectual Property Organization Report on AI and IP Policy, 2023

Statutes and Regulations:

35 U.S.C. § 101 (United States)
Arbitration and Conciliation Act, 2015 (India)
Copyright Act, 1957 (India)
Copyright Act of 1976, 17 U.S.C. § 102 (United States)
European Patent Convention, Article 52
Income Tax Act, 1961
Indian Contract Act, 1872
Lanham Act, 15 U.S.C. § 1051 (United States)
Patents Act, 1970 (India)
Patents Act 1977 (United Kingdom)
Trade Marks Act, 1999 (India)
TRIPS Agreement, Articles 27-34