AI Generated Art Ownership: Legal Framework, Key Cases, and Policy Insights

Artificial intelligence-generated art includes images, music, and text created with minimal human input using AI systems, algorithms, and vast datasets. Popular image generators like DALL-E 2, Stable Diffusion, and text-to-image tools such as Midjourney and ChatGPT have accelerated the creation of AI-generated content. As these AI models become more powerful and accessible, questions of ownership, authorship, and control over original works have become increasingly urgent.

Legal frameworks lag behind, creating uncertainty around AI-generated art ownership, originality, and fair use. For scholars and professionals, this is not just theoretical; it has real consequences for intellectual property (IP), the creative process, and global policy shaping the future of artistic works.

Legal Foundations of Ownership

Ownership of AI-generated art hinges on how existing IP laws (especially copyright, patents, and international agreements) interpret human authorship in works created by or with the help of machine learning.

Copyright and Authorship

Under U.S. copyright law, protection applies only to creative works made by human authors, as outlined according to the U.S. Copyright Office’s three-part Report on Copyright and Artificial Intelligence.^[1] In response to the AI-generated work A Recent Entrance to Paradise (2022), the U.S. Copyright Office rejected registration for an image created by an AI system, citing lack of input from a human being.^[2],[3] This sets a precedent for how OpenAI tools and others are treated under U.S. law.

Courts in the U.S. and U.K. have reinforced the need for a human author to claim rights over AI-generated content. Other jurisdictions offer different approaches:

• The U.K. grants limited-term copyright protection to the person “making arrangements” for computer-generated works under Section 9(3) of the Copyright, Designs and Patents Act.^[4]
• The EU maintains a human-centric definition of authorship, although policy reform discussions are underway.^[5]

Patents and Trademarks

Current patent law does not recognize AI models as inventors. In 2022, U.S. courts rejected applications naming AI as the sole inventor.^[6] For example, content generated by systems like Stable Diffusion or trained models built on large datasets cannot be patented unless a human is identified as the inventor.

In trademark law, humans can register AI-assisted logos if they show distinctiveness and use in commerce. However, widespread use of AI-generated designs challenges the traditional standards of originality, authorship, and source identification.^[7]

International Treaties

Global IP treaties like the Berne Convention and TRIPS don’t account for AI.^[8],[9] Recent EU initiatives focus on transparency and human authorship but stop short of recognizing generative AI as content creators.^[5] Cross-border differences mean copyright ownership, as well as copyright infringement and enforcement, vary by jurisdiction.

Key Cases Shaping the Landscape

Warhol Foundation v. Goldsmith

This 2023 Supreme Court case centered on whether Warhol’s adaptation of a photo was “transformative” under fair use.^[10] The Court ruled against Warhol, emphasizing commercial overlap and limited creative change. The ruling signals that AI-generated adaptations of copyrighted works may not qualify for fair use unless they serve a distinctly new purpose and reflect meaningful human input.

Zarya of the Dawn

The U.S. Copyright Office initially registered, then revoked, copyright for a graphic novel with AI-generated images, entitled Zarya of the Dawn.^[11],[12] While human-written text remained protected, images made using Midjourney were deemed too automated. Selecting outputs or providing text prompts was not enough to qualify as authorship; courts require significant human creative input.

These cases and similar rulings affirm that AI-generated works need demonstrable human contributions to qualify for protection.

Economic and Policy Implications

Incentives and Risk

Clear ownership rules promote innovation by giving artists, developers, and investors confidence. When legal rights are ambiguous, creators may hesitate to publish or invest, fearing infringement or revenue loss.

Defined rights help:

• Encourage artists to experiment with AI tools.
• Attract investment in AI development.
• Enable more accurate risk assessment in creative ventures.

Lack of clarity can deter innovation, as seen in markets where rights are contested or undervalued.

Licensing and Market Dynamics

The AI-generated artwork boom has spurred new licensing models:

• Standard licenses grant usage rights even without settled ownership.
• Custom contracts clarify control between creators, developers, and platforms.
• Royalty-sharing agreements divide income across contributors, such as prompt writers and model developers.

Buyers value clarity; art tied to enforceable rights commands higher prices. In contrast, uncertainty undermines trust and devalues work.

Legal Reforms and Best Practices

To keep pace with evolving technology, lawmakers and creators are proposing reforms that emphasize attribution, compensation, and registration.

Attribution and Compensation

New models aim to credit both human and AI developers:

• Multi-party attribution mirrors systems in film and music.
• Royalty agreements specify revenue splits.
• Documentation requirements log human input for legal clarity.

These steps support fairness and reduce litigation risk.

Registration Improvements

Even if AI works don’t fully qualify for copyright or patents, reforms could improve registration processes:

• Disclosure requirements clarify AI involvement.
• Modular forms distinguish between human and machine roles.
• Hybrid protection could offer partial rights based on human contribution.

Such practices would help build a reliable, transparent rights system for AI-assisted work.

Future Outlook and Debates

AI as Tool vs. Co-Creator

A key legal and philosophical divide is whether AI is a tool or co-creator:

• The tool model treats AI like a brush or camera; rights go to the human operator.
• The co-creator model suggests shared or no rights, especially when AI operates autonomously with little human direction.

This distinction affects whether courts grant exclusive rights to users or consider broader frameworks involving AI developers or exclude protection entirely.

Legislative Developments

Governments worldwide are exploring new laws:

• U.S.: Congressional hearings and Copyright Office consultations are ongoing, but no law yet defines AI-generated authorship.^[1]
• EU: The EU AI Act and copyright directives are influencing harmonized approaches with transparency and attribution standards.^[5]
• UK: The UK is reviewing its copyright frameworks, weighing whether to extend or limit rights for computer-generated works.^[4]

Conclusion

AI-generated art ownership challenges long-standing principles of intellectual property rights. Courts and policymakers are working to adapt, but legal clarity remains elusive. Without clear authorship standards, investment, innovation, and fair attribution are at risk.

Moving forward, scholars and lawmakers must collaborate to create workable, fair frameworks that:

• Recognize human contribution in AI-assisted works.
• Enable enforceable licensing and registration systems.
• Balance innovation with economic and legal certainty.

As new technologies and court rulings emerge, the field will continue evolving, requiring constant engagement from creators, legal experts, and policymakers alike to resolve copyright issues and encourage innovation.

[1] U.S. Copyright Office, Copyright and Artificial Intelligence, U.S. Copyright Office, https://www.copyright.gov/ai/.

[2] A Recent Entrance to Paradise (Creativity Machine, 2022).

[3] U.S. Copyright Off., Re: Second Request for Reconsideration for Refusal to Register A Recent Entrance to Paradise (Correspondence ID 1-3ZPC6C3) (Feb. 14, 2022), https://www.copyright.gov/rulings-filings/review-board/docs/a-recent-entrance-to-paradise.pdf.

[4] Copyright, Designs and Patents Act 1988, c. 48, § 9(3) (UK).

[5] Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act), 2024 O.J. (L 1689) 1.

[6] Thaler v. Vidal, 43 F.4th 1207 (Fed. Cir. 2022).

[7] Von Bryant, AI-Generated Logos Require Careful Steps to Protect IP Rights (June 4, 2025), https://news.bloomberglaw.com/us-law-week/ai-generated-logos-require-careful-steps-to-protect-ip-rights.

[8] Berne Convention for the Protection of Literary and Artistic Works, Sept. 9, 1886, as revised at Paris on July 24, 1971 and amended in 1979.

[9] Agreement on Trade-Related Aspects of Intellectual Property Rights, Apr. 15, 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1C, 1869 U.N.T.S. 299, 33 I.L.M. 1197 (1994).

[10] Andy Warhol Foundation for Visual Arts, Inc. v. Goldsmith, 598 U.S. 508 (2023).

[11] Kristina Kashtanova, Zarya of the Dawn (2022).

[12] Letter from U.S. Copyright Off. to Van Lindberg (Feb. 21, 2023), https://business.cch.com/ipld/CopyrightOfficeLetterLindbergZarya20230221.pdf.