The U.S. Copyright Office is presently undertaking a comprehensive examination of the application of copyright law to artificial intelligence (AI). In a recent conversation, a legal professional from the Office disclosed that it is actively registering creative works that include AI, contingent on certain stipulations.

In an article from April published in WIPO Magazine, Jalyce Mangum, Assistant General Counsel at the U.S. Copyright Office, mentioned that the agency has recorded over 1,000 works in which applicants revealed and disavowed AI-generated content, in line with the Office’s formal guidance. Mangum pointed out that the main factor to consider is whether AI is simply aiding human creativity or if it serves as the primary origin of the creative output.

A representative for the Copyright Office indicated that a report published in January 2025 revealed that hundreds of AI-enhanced works had already been registered, and that number has now surged to over 1,000.

This progression occurs amidst rising unease and pushback from artists and creators who are against the utilization of AI in creative fields. Many have expressed strong disapproval of AI-generated content in films and various media formats. Significant disputes have arisen regarding the use of AI in productions such as The Brutalist and Late Night With the Devil, the digital revival of Agatha Christie’s likeness, the AI-augmented restoration of The Wizard of Oz for The Sphere, and the widespread Studio Ghibli-inspired images produced using ChatGPT.

The Copyright Office has previously clarified that works created solely by AI are not qualified for copyright protection unless they are part of a larger creation that is predominantly the result of human authorship. This implies that images or text generated entirely by means such as DALL-E, ChatGPT, or Midjourney cannot receive copyright, irrespective of the effort invested in developing the prompts.

The Office’s complete guidance on AI-assisted works can be found on its website and was elaborated in the second segment of its report on copyright and artificial intelligence. This report reinforced that copyright protection is not applicable to works that are completely AI-generated or where human contribution to the creative process is negligible.

Nonetheless, works that are augmented by AI—rather than fully created by it—may be eligible for copyright. The report provides examples of permissible AI assistance, such as removing crowds from images, stabilizing video recordings, or employing ray tracing in visual effects.

The upcoming third segment of the Copyright Office’s report, anticipated later this year, will tackle the training of AI models and could have extensive effects on the future intersection of AI and copyright law.