On January 29, 2025, the U.S. Copyright Office released Part 2 of its report on Copyright and Artificial Intelligence. This section primarily focuses on the copyrightability of works created using generative AI. Part 3, which will address issues of copyright infringement and fair use involving generative AI, is highly anticipated.
Here are two key takeaways from Part 2:
1. No Change to the “Human Authorship” Requirement
In line with its stance in recent rulings and cases, the Copyright Office reaffirmed that human control over creative expression is required for copyright registration of generative AI outputs. The Office emphasized that user prompts (instructions given to the AI) are insufficient to demonstrate the level of human control needed for copyright eligibility. As a result, even extensive prompt engineering will not result in copyrightable work using current generative AI technology. The key passage from the Office reads:
“In theory, AI systems could someday allow users to exert so much control over how their expression is reflected in an output that the system’s contribution would become [protectable]. The evidence as to the operation of today’s AI systems indicates that this is not currently the case. Prompts do not appear to adequately determine the expressive elements produced, or control how the system translates them into an output.”
The Copyright Office reaffirmed that copyright protection is available for:
- (a) Human-created works used as inputs/prompts that are perceptible in the AI-generated output;
- (b) Creative selection, coordination, or arrangement of material in the output (i.e., compilations);
- (c) Creative modifications of the output; and
- (d) The prompts themselves, provided they are sufficiently creative (though not the outputs generated in response to the prompts).
2. Foreign Laws Mostly Align with U.S. Positions
The Office also conducted a comparative analysis of the copyrightability of AI-generated works in countries including South Korea, Japan, China, the EU, the UK, Hong Kong, India, New Zealand, Canada, and Australia. It concluded that most of these nations “that have addressed this issue so far have agreed that copyright requires human authorship.”
An interesting note from the report was the Office’s reference to a 2023 decision by the Beijing Internet Court, which granted copyright protection for an AI-generated image in an infringement case. The Court determined that the “selection of over 150 prompts combined with subsequent adjustments and modifications demonstrated that the image was the result of the author’s ‘intellectual achievements,’ reflecting his personalized expression.” Given the Copyright Office’s position on prompts, it remains unclear whether this decision aligns with the U.S. view—though it may be partially consistent concerning the “adjustments and modifications.” Some commentators have cited this case as an example of China diverging from the U.S. approach by allowing copyright protection for AI-generated images.
The Office also pointed out that legal stances on AI-generated works are evolving in many countries, and in some cases, remain unclear.
As we await Part 3 of the report, our DC IP lawyers see it’s clear that questions surrounding AI copyright infringement and fair use are heating up, especially as numerous AI copyright infringement cases are moving through U.S. courts. Preliminary answers may soon be on the horizon.