A copyright report left AI authorship a case-by-case question
The Copyright Office's second AI report concludes existing law needs no change, but sets no fixed threshold for human contribution.
- Historical event
- January 29, 2025
- First source published
- January 29, 2025
- Site publication
- September 18, 2026
The record
On 29 January 2025 the United States Copyright Office released Copyright and Artificial Intelligence, Part 2: Copyrightability, the second installment of a study opened by the office's August 2023 Notice of Inquiry. The report itself states it received more than 10,000 comments in response to that inquiry, roughly half of them addressing copyrightability, and that it "analyzes the type and level of human contribution sufficient to bring" AI-system outputs within copyright's scope. The office's own AI initiative page, retrieved 16 September 2026, situates Part 2 between an earlier July 2024 report on digital replicas and a pre-publication Part 3 on generative-AI training released in May 2025, with a final version of that third part still described there as forthcoming.
What the documents show
The report's executive summary sets out its conclusions plainly: existing copyright law can resolve AI questions without legislative change; using AI as an assistive tool does not defeat copyright in the resulting work; purely AI-generated material, or material lacking sufficient human control over its expressive elements, receives no protection; and, based on how generally available systems function today, prompts alone do not supply that control. The report also states that a human author retains copyright in whatever of their own expression is perceptible in an AI-assisted output, including creative selection, arrangement or modification. What the report does not supply is a bright line for how much involvement is enough; it says the question must be resolved case by case, leaving individual registration decisions to fill in the practical detail.
Why it stays interesting
This is an editorial note: the report matters less for any single rule than for what it declines to settle. By keeping copyrightability a case-by-case inquiry rather than fixing a threshold, the office leaves authors, publishers and platforms negotiating disclosure and contract language around a boundary that moves with each registration decision rather than a fixed one. A reader assessing a book's front matter, a platform's content policy, or a publisher's proposed clause is, in effect, applying this report's still-open middle ground to one particular case, and should expect no single citation to close the question for them.
What to read next
The office's Part 1 report on digital replicas addresses a related but distinct question, a person's voice or likeness rather than authorship of a work, and repays reading against Part 2's narrower focus. A reader might also compare this report's case-by-case approach with how the office resolved one concrete dispute over a graphic novel's AI-generated panels.
- Which elements of a given work does its creator claim as their own selection or arrangement, and which came from the tool unmodified?
- Does a publisher's contract or a platform's disclosure policy track the report's line between assistive use and purely generated output?
- What would change in this analysis once the office's training report, still pending in final form, addresses the inputs used to build these systems?
The report closes one chapter of a longer study without pretending to close the underlying question, and a careful reader will learn more from what it declines to fix than from the conclusions it draws.
Sources & reading trail
The report's own executive summary and conclusions on human authorship and AI-generated content.
Source published: 29 January 2025 · Retrieved: 16 September 2026
Confirms the 29 January 2025 publication date and the report's place among the initiative's three parts.
Source published: Not established · Retrieved: 16 September 2026
Reports, filings, records and studies establish the entry; the reading of why it stays interesting is Chapter Ledger editorial analysis. This retrospective draft does not imply the site published on the event date.
Continue reading
- A district court denied copyright to a machine-made image
- A comic's text kept copyright while its AI images lost it
- Browse the complete the ledger archive
Sources & reading trail
- Copyright and Artificial Intelligence, Part 2: Copyrightability
Source published: January 29, 2025 · Retrieved: September 16, 2026 - Copyright and Artificial Intelligence (AI initiative overview)
Retrieved: September 16, 2026
The documents above establish the record. The reading and the questions are this publication’s editorial analysis, written after the fact.
Published September 18, 2026, not on the date of the event described.