Introduction

The rapid development of artificial intelligence systems has unsettled the foundations of intellectual property law by challenging long-held assumptions about human authorship and inventiveness. AI can autonomously compose music, draft literature, design logos, invent technical solutions, and even co-author scientific papers. Yet, global IP regimes were designed around the notion of a human creator or inventor. Questions arise:

• Who owns AI-generated inventions and creative works?
• Are AI outputs eligible for legal protection?
• How can the global IP system adapt while supporting innovation and balancing public interest?
• What ethical and transparency standards must be met?

These questions cut across all IP domains, from copyright and patents to trademarks and trade secrets.

AI and Copyright Law

1. Copyrightability of AI-Generated Works

A critical legal issue is whether AI-generated content—text, music, images, and other outputs—qualifies for copyright protection. The consensus in major jurisdictions is that copyright attaches only to works with human authorship. The U.S. Copyright Office and courts have repeatedly ruled that AI-generated materials, produced without human creative input, are ineligible for copyright registration^[1][2][3]. The famous Thaler case (2023 U.S. District Court, Washington D.C.) and similar rulings worldwide reinforce this principle^[2][3].

Table: International Copyright Criteria for AI Outputs (2025)

AI systems trained on existing copyrighted works have also sparked litigation—creators claim infringement if their works are used without permission in model training^[3][5].

2. Data Mining, Fair Use, and Transparency

Countries are grappling with how "fair use" or "text and data mining" (TDM) exceptions apply to AI development. The EU now mandates that developers disclose training data sources, while countries like China and France have started restricting unlicensed data mining^[5]. Canada, India, and the U.S. are considering new legislation to clarify these rules^[4][5].

AI and Patent Law

1. Inventorship and Patentability

Patent law universally requires a human inventor. Landmark cases such as the DABUS litigation (UK, Europe, U.S.) held that AI itself cannot be named as an inventor or qualify for patents^[1][6][7]. Instead, rights are usually assigned to those who develop, own, or oversee the AI that produced the invention^[7].

[image:1]

Above: Inventorship in AI patents remains tied to human identification.

However, this creates ambiguity when inventions are primarily conceived by autonomous AI systems. Practically, it hampers patenting genuinely novel AI-generated innovations and could stifle socially beneficial technologies if reforms are not enacted^[6][7].

2. Risks of "Patent Thickets" and Bias

AI-driven automation in patent filings has led to exponential growth—AI-related patent applications have increased by 33% since 2018 and span 60% of technology subclasses^[8]. This raises concerns about "patent thickets," where numerous overlapping patents block competitors and innovation^[7].

AI also reflects and can perpetuate historical biases present in earlier patent data, risking unfair or uneven access to protection for minority inventors or technologies^[7].

Trademarks, Trade Secrets, and Other IP Issues

• Trademarks: AI-generated logos or brand names create ambiguity about rightful ownership and originality.
• Trade Secrets: AI systems increase risk of data leaks and require robust protocols for protecting proprietary algorithms or training data.
• Dual Role of AI: AI acts as both a threat (content generation, counterfeiting) and a tool for enforcement (monitoring, detection, automated rights management)^[9].

Emerging Policy and Legislative Responses

• The U.S., Germany, India, China, EU, and others are reviewing IP statutes and have launched public consultations or committees to address the AI–IPR nexus^[4][10].
• In India, a high-level committee was recently formed to recommend copyright reforms for AI-generated works^[4].
• Policymakers are considering new standards for originality, inventorship, and liability in the context of AI^[11][7].

Graph: Growth in AI-related Patent Filings (2018–2025)

[image:2]

Litigation Trends

2025 has seen a flurry of lawsuits involving major technology firms and AI companies over copyright infringement, authorship, and fair use in model training^[3][5]. Decisions are beginning to converge on key principles:

• Human creativity is essential for copyright.
• AI-generated work derived from or copying protected materials may infringe copyright.
• Reform is needed for greater legal clarity^[2][3][5].

Comparative Perspectives

Ethical and Social Implications

• Transparency: Clearer disclosures on how AI is created and trained are needed to foster trust and curb infringement risks^[7].
• Balance: Legal systems must balance incentives for innovation with creators’ rights and the public interest.
• Bias and Fairness: Addressing bias in patent and copyright systems remains a central ethical challenge^[7].

Conclusions and Future Directions

AI is testing the boundaries of intellectual property law at every turn. While legal frameworks worldwide have begun adapting, more harmonized reforms and forward-looking policies are urgently needed. Ensuring protection for genuine human creativity, fostering responsible AI-led innovation, and clarifying accountability and ownership are essential for the future of IPR in the age of AI.