Ipr In Ip Audits Of Creative Ai Technologies.
IPR IN IP AUDITS OF CREATIVE AI TECHNOLOGIES
1. Introduction
Intellectual Property (IP) audits are systematic reviews of a company’s intellectual property assets, ensuring they are protected, well-managed, and monetized effectively. In the context of creative AI technologies, such as AI-generated art, music, and content creation tools, IP audits become complex due to questions about ownership, patentability, copyright, and licensing.
Creative AI technologies encompass a range of tools and algorithms that enable machines to autonomously generate content in various fields (e.g., visual art, music, writing, design). With the rise of AI systems like DeepDream, OpenAI’s GPT-3, DALL·E, and others, the question arises:
Who owns the content generated by AI?
How can AI tools and their outputs be licensed or monetized?
What IP protections should be applied to AI-driven creative works?
An IP audit of AI technologies evaluates these rights and assets, ensuring that:
Ownership of AI-generated content is clearly established.
Patents for novel AI processes or systems are appropriately filed.
Copyrights for the content produced by AI systems are attributed correctly.
Licensing terms for AI-driven creations are set to avoid disputes.
2. Key IP Issues in Auditing Creative AI Technologies
Ownership of AI-Generated Content
Who owns the rights to AI-generated art, music, or writings? The owner could be the creator of the AI system, the user who interacts with the AI, or potentially even the AI system itself (if legally recognized as a creator).
Patents on AI Algorithms or Tools
AI-driven tools or technologies that improve content generation (e.g., better image rendering techniques, novel algorithms) can be patented. It’s essential to perform an audit of any innovative AI algorithms to see if patent protection is needed.
Copyright in AI-Generated Works
If an AI generates a novel image or song, should it be protected by copyright? The question of whether AI-generated works are eligible for copyright without human involvement is still developing in most jurisdictions.
Trade Secrets and Confidentiality
Proprietary AI models and datasets that drive creative tools are often protected as trade secrets. IP audits ensure that these trade secrets are safeguarded and appropriately licensed.
Licensing and Royalty Management
Ensuring that licensing agreements for AI-generated works are clear, enforceable, and protect IP rights. This includes understanding the use of third-party datasets, the application of open-source licenses, and the commercial exploitation of AI tools.
CASE LAWS (DETAILED ANALYSIS)
Case 1: Naruto v. Slater (2018, US)
Facts:
A monkey named Naruto took selfies with a photographer’s camera.
The photographer filed for copyright ownership, but Naruto was the "creator."
Legal Issue:
Can non-human entities (in this case, an animal) claim ownership over creative works?
Judgment:
The court ruled that only humans could hold copyright over creative works.
Therefore, Naruto, the monkey, was not recognized as the author of the photographs.
Relevance:
This case is particularly relevant to AI-generated content. It highlights the issue of whether AI, a non-human entity, can own a copyright for its generated works.
AI-generated assets may face similar challenges, and ownership of AI-created content will likely belong to the developer or user interacting with the AI, rather than the AI itself.
During an IP audit, the authorship of AI-generated works must be clarified to ensure proper copyright attribution.
Case 2: Feist Publications, Inc. v. Rural Telephone Service Co. (1991, US)
Facts:
Feist Publications copied data from a telephone directory published by Rural Telephone and created a competing directory.
Rural Telephone argued that its selection and arrangement of data was copyrightable.
Legal Issue:
Can the selection and arrangement of facts be copyrighted?
Judgment:
The court ruled that facts are not copyrightable, but the arrangement or creative selection can be if there is sufficient originality.
Relevance:
This case is particularly relevant for AI systems that use large datasets to generate creative outputs.
In an IP audit of AI tools, it is crucial to assess whether the AI’s use of a training dataset is copyright-compliant, particularly if the AI is generating derivative works based on datasets that may be public domain or licensed.
The case underscores the importance of originality in both the use of datasets and the generated content.
Case 3: Thales Visionix, Inc. v. United States (2017, US)
Facts:
Thales Visionix developed a motion-tracking system used in virtual reality and related technologies.
The U.S. government argued the patent was an abstract idea and not patentable.
Legal Issue:
Can an abstract idea implemented in software be patented?
Judgment:
The court ruled that abstract ideas implemented via software cannot be patented unless they contribute to a technical improvement.
Relevance:
This case is crucial for AI patentability. When conducting an IP audit, AI systems or algorithms should be evaluated for their novelty and technical advancement.
If the AI process improves upon an existing method (e.g., rendering or interaction design), it may be eligible for patent protection. AI technologies that simply automate an existing process without technical advancement are not patentable.
Case 4: Oracle America v. Google Inc. (2011, US)
Facts:
Oracle sued Google for using Java in its Android operating system without proper licensing.
The case involved the use of Java API code in Android, which Oracle argued was protected by copyright.
Legal Issue:
Can the use of Java's API code in a software system be considered fair use under copyright law?
Judgment:
The court ultimately ruled that Oracle was entitled to damages for using Oracle’s copyrighted code without permission.
However, Google successfully argued fair use on the basis of transformative use and the commercial nature of the use.
Relevance:
This case has significant implications for the licensing of AI technologies that rely on third-party datasets or APIs to create content.
During an IP audit of AI technology, companies must assess whether third-party datasets are being used in compliance with licensing agreements, and if AI-generated outputs based on those datasets could be considered derivative works.
The case also highlights the importance of license management and ensuring fair use claims are defensible.
Case 5: Warhol Foundation v. Lynn Goldsmith (2019, US)
Facts:
The Warhol Foundation used Lynn Goldsmith’s iconic photograph of Prince to create a series of artworks.
Goldsmith sued, arguing Warhol’s work was not sufficiently transformative and infringed on her copyright.
Judgment:
The court ruled that Warhol’s use was not transformative enough to qualify for fair use.
The court upheld Goldsmith's copyright.
Relevance:
This case underscores the importance of transformative use in the context of derivative works. For AI-generated content that is based on or inspired by existing works, determining whether the output is sufficiently transformative is critical to avoid copyright infringement.
In an IP audit, when reviewing AI-generated art, companies must assess whether fair use or transformative use applies, especially if AI is trained on or generates works inspired by copyrighted material.
Case 6: Hermès International v. Rothschild (2022, US)
Facts:
Rothschild created MetaBirkin NFTs resembling Hermès' famous Birkin bags.
Hermès sued Rothschild for trademark infringement over the unauthorized use of its brand.
Judgment:
The court ruled in favor of Hermès, asserting that trademark infringement occurred even in the NFT space.
Relevance:
AI-generated assets, especially in the context of NFTs, can infringe on trademarks if they mimic real-world brands.
IP audits for creative AI technologies must include assessments of how AI-generated assets relate to existing brands or trademarks to avoid potential infringement.
4. Conclusion
IP audits of AI technologies must be comprehensive, addressing key issues such as:
Ownership of AI-generated content (who holds the copyright, trademark, or patent?)
Licensing models (non-exclusive, exclusive, and royalty-based licensing)
Patentability of innovative AI algorithms
Copyright compliance for AI-generated works and their derivative nature
Trade secrets protection for proprietary AI models or datasets
Case law plays an essential role in shaping the landscape for IP audits, particularly as AI continues to generate content and innovate across industries. As AI-generated assets become more common, audits will increasingly need to navigate authorship and ownership rights, ensuring proper license management and compliance.
RELATED Blog
comments