The advancement and integration of artificial intelligence (AI) technology, particularly generative AI tools such as ChatGPT and Stable Diffusion, have revolutionized content creation. However, rapid evolution and widespread use has raised significant legal and ethical questions in the field of intellectual property (IP), particularly regarding the copyright status of AI-generated content (AIGC). Key concerns include whether AIGC is copyrightable, who qualifies as its author, who owns its copyright, and whether AI service providers can be held liable for copyright infringement. McCoy Russell explores these concerns along with recent guidance from the U.S. Copyright Office (USCO). We recognize that different countries are facing the AI copyright question with different results.
AI-Generated Content: Can It Be Copyrighted?
The U.S. Copyright Office has taken a stance: AI-generated works are not eligible for copyright protection unless there is evident human creative input. The use of AI to assist in creation or inclusion of AI-generated material does not bar copyright protection. As such, when an individual applies for copyright on a work containing AI-generated elements, it may be required to identify which portions were created by AI and acknowledge that those sections are not subject to copyright protection.
Who Is the Author of AI-Generated Content?
Copyright law traditionally requires human authorship, and the USCO maintains that a human must have played a substantive creative role in the work’s production. The agency emphasizes that human involvement, such as selecting, arranging, or modifying AI-generated content, is necessary for copyright eligibility. Without such input, the work remains uncopyrightable.
Who Owns the Copyright to AI-Generated Content?
Under U.S. law (17 U.S. Code §201), copyright automatically belongs to the author upon the work’s creation. However, courts have consistently ruled that copyright requires human authorship. This principle was reinforced in the 2018 “monkey selfie” case, Naruto v. Slater, where a court ruled that a non-human entity could not own copyright. Similarly, in Thaler v. Perlmutter (2023), the U.S. District Court for the District of Columbia rejected a claim for copyright in an AI-generated work that had no human involvement. These rulings, combined with guidance from the USCO, clarify that neither AI developers, AI tools themselves, nor users providing prompts can claim ownership over AI-generated works. As a result, it is considered that AIGC may arguably be considered part of the public domain and not protected by copyright.
Are AI Service Providers Liable for Copyright Infringement?
The July 2024 Copyright and AI report provides recommendations on liability concerning AI-generated content. It suggests that AI service providers should not be held liable for creating unauthorized digital replicas but could face secondary liability for distributing them. The report proposes a safe harbor defense for online service providers who promptly remove unauthorized replicas upon notification. Additionally, it supports a licensing framework that allows individuals to monetize digital replica rights, ensuring a balance between copyright enforcement and free speech considerations. Of note are comments where the report does not advocate for federal preemption of state laws but instead recommends a federal “floor” that states can build upon to enhance protections.
The USCO is expected to release further guidance on the intersection of AI and copyright throughout 2025. Legal and regulatory processes will need to adapt to address the growing complexities of AI-driven content creation. McCoy Russell is actively monitoring these developments in the US and in other jurisdictions and will continue to provide insights into the shifting landscape to support the development of its clients IP portfolios.
Recent Comments