When 600 Prompts Still Aren't Enough: What Allen Vs. Perlmutter Means For Ownership, Copyright, And Creative Contracts

Who owns creative work produced by AI? This has become a common question in litigation and the U.S. Copyright Office continues to answer the same way: not the person who merely prompts the system (no matter how many prompts are used). The Case: Allen v. Perlmutter. Jason Allen created an image titled Théâtre D'opéra Spatial […]

Who owns creative work produced by AI? This has become a common question in litigation and the U.S. Copyright Office continues to answer the same way: not the person who merely prompts the system (no matter how many prompts are used).

The Case: Allen v. Perlmutter. Jason Allen created an image titled Théâtre D'opéra Spatial using the AI platform Midjourney. After refining the output through more than 600 prompts, the image won first place at the Colorado State Fair. Allen then sought copyright registration. The U.S. Copyright office refused and Allen sued.

Allen's Argument. Allen argued that 600 prompts would certainly meet the originality threshold under copyright law and that using AI was no different than using a camera or other creative technology (citing the Burrow-Giles Lithographic Co. v. Sarony, a landmark photography case).

The Copyright Office's Position. In its motion for summary judgment, the Copyright Office disagreed, reaffirming its long-standing position that copyright protects human-authored expressive elements only; prompts are ideas or instructions, not authorship; the actual AI system (and not the user) determines the final expressive output; and time, effort, and iteration do not substitute for human creation.

The Office also emphasized that Allen could have sought a limited registration covering only his post-generation edits, while disclaiming the AI-generated portions.

AI, Authorship, and Ownership: A Legal Inflection Point

Generative AI has become a core tool for brands, agencies, creators, and production studios. But this case highlights a growing legal disconnect between how creative work is made and how copyright law assigns ownership. While Allen v. Perlmutter is still pending, the Copyright Office's position carries real-world consequences for businesses using AI in creative workflows.

1. Prompting Does Not Equal Ownership. No matter how detailed, original, funny or iterative, prompting alone is unlikely to establish copyrightable authorship under current law. That reality undermines assumptions many businesses make about owning AI-generated assets.
2. Copyright Gaps Create Business Risk. Without copyright protection, companies may lose out on crucial enforcement rights such as preventing copying by competitors, losing exclusivity in brand assets or marketing materials; facing challenges enforcing licenses or assignments; and seeing reduced asset value in acquisitions or investments. All of these rights (or lack thereof) directly affect monetization and brand protection.
3. Hybrid AI Works Require Precision. The Copyright Office has made clear that authorship is not all or nothing. Human contributions—such as editing, selection, arrangement, or modification—may be protectable, but only if properly identified and disclaimed. Failure to address this correctly can result in no registration at all.
4. Emerging Technology. AI is still a new and emerging technology. Unlike cameras or creative services software, generative AI autonomously determines expressive elements. Until courts or Congress say otherwise that distinction will continue to limit how far traditional copyright analogies go.

The Bigger Picture: Contracts Now Carry the Weight. As courts and regulators work through these issues, ownership certainty for AI-assisted creative work now lives primarily in contracts—not copyright law. No party in the creative ecosystem including brands, agencies, creators, and platforms using generative AI should rely on default assumptions about ownership, work-for-hire status, or enforceability.

What Creative Businesses Should Do Now. Companies using AI for advertising and marketing content, brand design and visual assets, influencer and creator campaigns, and even film, music and digital production should ensure that their agreements address these big issues. Contracts should outline ownership and licensing of AI-generated outputs, allocation of risk for non-copyrightable material, warranties and indemnities tied to AI use and disclosure obligations. On top of that, don't forget compliance with the copyright regulations and most current FTC guidance.

Protect Your Creative Assets Before Disputes Arise. If your organization uses AI in creative work, now is the time to get your contracts right, evaluate your AI workflows and draft agreements that protect your creative and commercial interests.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.