QIM Score: 78/100 — published under the house rule: no post goes live unscored.
Routes: Blue Data Law · Digital Rights
The Data Hook
In January 2025, the U.S. Copyright Office published the second part of its artificial-intelligence report and settled, for now, the question every brand using generative imagery should have been asking: who owns the picture the machine made? The answer, delivered in measured bureaucratic prose, is brutal in its implications — mostly, nobody.
Prompts Are Requests, Not Authorship
The report reaffirms that human authorship is the bedrock of copyright, and that works generated entirely by AI are not copyrightable. Crucially, it rejects the industry's favorite workaround: detailed prompting. Given currently available technology, the Office concluded, prompts alone do not provide sufficient human control to make the user the author — a prompt is an instruction to a system that fills the expressive gaps itself, however elaborate the instruction. Refining a prompt a hundred times is, in the Office's analysis, a hundred requests, not an act of authorship. The picture that emerges belongs to no one; it arrives, in effect, in the public domain.
Where Human Ownership Survives
Three doors stay open. First, perceptible human expression within the output: if your hand-drawn art, your photograph, or your written text is incorporated and remains visible, your contribution stays yours. Second, creative selection, coordination, and arrangement: a human who curates, sequences, and assembles AI elements into a larger work owns the compilation — the arrangement, though not the underlying images. Third, material modification: substantive human editing of an output can yield protectable expression in the modified result. The pattern across all three is the same — copyright follows the human hand, wherever it actually touched the work.
The Strategy This Dictates for Brands
If the images can't be owned, stop building the moat out of images. Own the names: trademarks protect brand and persona names regardless of how the artwork was made. Own the look as trade dress: a consistent visual system — palette, composition, the recurring character — functions as a source identifier the Lanham Act protects even when no single picture is copyrightable. Own the arrangement: register catalogs, lookbooks, and campaigns as compilations, because the scraper who takes your curated set takes your protected selection. Own the words: captions, scripts, character bibles, and posts are human-authored and registrable. And skin every published image with human-authored elements — wordmark, caption, design frame — so that copying the image as it actually circulates means copying something you own.
Questions Creators Keep Asking
If I inpaint, retouch, and composite the AI output extensively, do I own the result?
You own what you contributed. Material human modification yields protection for the modified expression — the composition you built, the elements you painted, the arrangement you imposed — while the untouched generated regions remain unprotected. Registration practice reflects this: the Office registers such works with AI-generated material disclaimed, which is far better than nothing because infringers rarely copy only the disclaimed parts.
Should I even bother registering anything?
Emphatically yes. Register the compilations (catalogs, lookbooks, campaign sets), the text (bibles, scripts, posts), and the human-modified works with appropriate disclaimers. Registration is the ticket to statutory damages and fees on the human-authored layers — and a registered compilation plus a strong trademark portfolio covers, in practice, nearly every real-world copying scenario a brand faces.
Could the rule change?
The Office itself flagged that its prompts-only conclusion reflects current technology, and tools offering granular, perceptible human control keep arriving. The trajectory favors more protectable human contribution over time — which argues for documenting your creative process now, so that yesterday's borderline work can be re-evaluated under tomorrow's standard with evidence in hand.
What to Do
The report didn't take anything from creators that they actually had; it clarified where the value always lived. Identity, language, arrangement, and consistency were the moat before generative tools and remain the moat after. Put the human hand where the law can see it, and put the trademark where the copier can't avoid it — then run an annual registration audit alongside your trademark renewals. A free Blue Data Law consult maps which layers of your brand are defensible and what to file first.
Blue Data Law — free consult | Michael Benavides, Esq., CA Bar No. 270714 | 707-362-4166 | attorneymichaelbenavides.com
ATTORNEY ADVERTISING. Blue Data Law is a trade name of the law practice of Michael Benavides, Esq., California State Bar No. 270714. General information only — not legal advice; no attorney-client relationship is formed by reading this. Authority cited is as of mid-2026 (U.S. Copyright Office, "Copyright and Artificial Intelligence," Part 2 on Copyrightability, Jan. 2025; Lanham Act trademark and trade-dress protection; Copyright Act human-authorship requirement) — verify current authority, as policy and registration practice evolve. Prior results do not guarantee a similar outcome.
