The Impact of AI on Intellectual Property Law

Introduction

The dawn of artificial intelligence (AI) marks one of the most profound technological shifts of  the 21st century, rapidly transforming the way we live, work, and create. As AI systems become  increasingly sophisticated, they now perform tasks such as generating music, designing new  molecules, drafting patent claims, and even creating works of art and literature; domains  historically governed by human ingenuity. This radical expansion in capability fundamentally  tests the boundaries and definitions long established in intellectual property (IP) law. Where  the law once presumed a clear human author or inventor, AI challenges foundational  assumptions: Can a machine be an author or inventor? Who owns the creations of autonomous  or semi-autonomous systems? How should disputes regarding ownership, attribution, and  infringement be resolved when non-human agents are involved?

Lawmakers, courts, and IP practitioners worldwide now face these unprecedented challenges.  The issues are not merely theoretical. In recent years, lawsuits concerning AI-generated  content, nuanced policy debates, and a surge in regulatory activity around the globe have  brought the impact of AI on IP rights to the fore. As AI permeates creative, scientific, and  commercial fields, it disrupts not only the ethical and economic balance of IP regimes but also  requires a reexamination of what property and creativity mean in a world where machines learn,  adapt, and generate on their own. Navigating these questions requires a careful balance:  promoting innovation and competitiveness while safeguarding the incentives and rights of both  human creators and rights holders in a rapidly evolving technological landscape. 

AI and Copyright: Creativity, Authorship, and Ownership

A central pillar of IP law is copyright, traditionally designed to incentivize and reward human  creativity. However, AI-generated works are often created with minimal or no human input which challenges the conventional notion that only humans can be authors. The U.S. Copyright  Office, for instance, has steadfastly maintained that “human authorship is a prerequisite for  copyright protection.” This stance was reaffirmed by a Washington D.C. district court in 2023, which denied copyright registration to works produced solely by AI, holding that only human  beings qualify as authors under U.S. law. The rationale is that copyright protects the expression  of ideas by humans, not ideas themselves or works generated autonomously by machines. 

The question becomes more complex with collaborative or assisted works, where AI tools are  used as creative aids. Here, courts and regulators must delineate how much human input  suffices for copyright eligibility. While works “merely assisted by AI” may remain protectable,  wholly AI-generated works do not enjoy copyright in the U.S. Other jurisdictions, such as the  European Union, have begun to contemplate specific digital rights, but no global standard  currently exists. 

There are also thorny issues around the use of copyrighted materials to train AI models.  Creators and rightsholders have raised concerns about unauthorized scraping and reproduction.  Some jurisdictions, notably the EU, now allow rights holders to object to the use of their works  for AI training, and high-profile lawsuits have been filed in the U.S. against major AI platforms,  with the likely outcome being shaped by future litigation or legislative action. 

Patents and AI: Inventorship, Novelty, and Prior Art

Patent law faces parallel dilemmas, particularly surrounding inventorship. Under most current  frameworks, an inventor must be a human being. This was put to the test by the widely  discussed “DABUS” case, where AI was listed as an inventor in patent applications filed  worldwide. Every major jurisdiction US, UK, EU, has refused patent applications without a  human inventor, citing statutory requirements. 

Challenges extend beyond inventorship. Assessing novelty and inventive step becomes  problematic with AI-generated inventions, as AI systems can recombine existing knowledge in  unpredictable ways. Moreover, the emergence of AI-generated prior art, AI tools autonomously  publishing technical disclosures may flood the patent landscape and complicate standards of  novelty and obviousness. Patent applicants now face the risk that AI-generated content could  surface as prior art, potentially undermining otherwise patentable inventions, especially in fast evolving fields like pharmaceuticals and electronics.

Additionally, questions arise around disclosure and replicability. AI inventions often rely on  complex algorithms and vast datasets, making it difficult to sufficiently disclose how an  invention works, a key requirement for patentability. Policymakers and courts may need to  provide guidance on how much algorithmic detail or data transparency is required to meet the  enablement standard in AI-related patents. 

Trade Secrets and AI: Confidentiality in a Transparent Age

Trade secret law offers another avenue for protecting AI innovations, particularly when patent  or copyright protection is unavailable due to human authorship requirements. Trade secrets  protect valuable, confidential business information such as algorithms, source code, and  proprietary datasets that derive economic value from secrecy and are subject to reasonable  efforts to maintain that secrecy.

However, AI assets’ intangible, dynamic, and distributed nature complicates trade secret  enforcement. AI systems frequently evolve, learn, and adapt, making it difficult to define or  identify the specific secret at issue. The growing demand for transparency in AI systems due  to regulatory and ethical imperatives, such as explainability in high-stakes applications can  clash with the imperative of secrecy. Furthermore, the global nature of AI development  increases the risk of cross-border misappropriation, where varying trade secret laws pose  enforcement hurdles. 

In practice, courts have allowed AI system designers and owners to claim trade secrets over  certain algorithms, even when the specifics of the algorithm may not be fully known or  understood due to the complexity of machine learning models. However, litigation in this area  is novel and evolving, requiring plaintiffs to adequately describe both the secret and efforts to  protect it, without disclosing too much and thereby forfeiting protection. 

Emerging Trends: Trademark, Design Rights, and the Future of IP in AI

While copyright, patent, and trade secret law dominate discourse, AI also impacts trademark  law and design rights. For example, AI-driven branding tools can generate logos and brand  names, raising questions about originality, distinctiveness, and ownership. Enforcement becomes equally tricky, as AI-enabled infringers can automate the creation of near-identical or  confusingly similar marks, complicating detection and legal action.

There is also debate around the adaptation of existing IP regimes or the creation of sui generis  (unique) protections for certain types of AI-generated works. Some suggest that, rather than  stretching current laws to accommodate machines, legislatures may need to introduce  specialized legal mechanisms to address the unique challenges and promote responsible  innovation.

Global harmonization, meanwhile, remains elusive. Different jurisdictions apply varying  standards for authorship, inventorship, protection, and infringement, creating uncertainty for  international companies and innovators. As AI technologies continue to advance, the  divergence may widen unless or until international organizations such as the World Intellectual  Property Organization (WIPO) or multinational legislation sets clearer boundaries.

Conclusion

The impact of AI on intellectual property law is transformative and deeply complex. As AI  becomes a participant, rather than merely a tool in creative and inventive processes, intellectual  property law is being pushed to its conceptual and practical limits. Regulators and courts must  strike a careful balance: protecting human creativity and labor, incentivizing innovation, and  promoting fair competition, while also recognizing the reality of machine-generated works.  Policymakers must grapple with issues of authorship, inventorship, transparency, and cross border enforcement, ensuring that IP law remains relevant and effective in the digital age.

Meaningful reform may require new legal categories or frameworks, designed thoughtfully to  support the twin goals of innovation and equity. Ultimately, the successful accommodation of  AI within the IP landscape will determine not only the contours of future legal regimes but also  the incentives for the next era of technological and cultural progress

Reference(S):  

1. U.S. Copyright Office Policy on AI Authorship

U.S. COPYRIGHT OFFICE, REGISTRATION GUIDANCE ON COPYRIGHT CLAIMS  INVOLVING ARTIFICIAL INTELLIGENCE (Dec. 2022),  https://www.copyright.gov/policy/artificial-intelligence/.

2. Recent D.C. District Court Case Denying Copyright to AI-Generated Works Thaler v. Perlmutter, No. 1:22-cv-00088 (D.D.C. May 20, 2023).
3. The “DABUS” Patent Cases (AI Inventorship Litigation)

In re Application of Stephen L. Thaler, 837 F. App’x 652 (Fed. Cir. 2021); see also United Kingdom Intellectual Property Office, Decision on AI Inventorship (2021),  https://www.gov.uk/government/publications/artificial-intelligence-and-inventorship; European Patent Office, EPO Board of Appeal Decision on AI Inventor (2022).

4. U.S. Patent Statute Mandating Inventors Must Be Human

35 U.S.C. § 100(f) (defining “inventor” as an individual).

5. European Union Digital Single Market Directive (on Copyright and AI) Directive 2019/790, 2019 O.J. (L 130) 92 (EU).
6. Trade Secret Protection under U.S. Law

Defend Trade Secrets Act of 2016, 18 U.S.C. §§ 1831–1839.

7. World Intellectual Property Organization (WIPO) AI and IP Issues Report World Intellectual Property Organization, “WIPO Technology Trends 2021: Artificial Intelligence” (2021), https://www.wipo.int/edocs/pubdocs/en/wipo_pub_1055.pdf.