Whose Prompt Is It Anyway? Navigating AI, Copyright Law, and Prompt Plagiarism

Whose Prompt Is It Anyway?: If you have spent any time in the genealogy community lately, you have noticed that AI prompts have become the new hot commodity. Prompt libraries, prompt packs, prompt optimization services. Everyone seems to be using them, sharing them, or arguing about whether someone stole theirs. The questions are piling up faster than unprocessed census images: Can you copyright a prompt? What happens when an AI rewrites your prompt for you? And if someone copies your carefully crafted instructions, is that theft or just the nature of the beast?

These are genuine legal and ethical questions that affect genealogists, family historians, content creators, and anyone building products around AI tools. So let’s dig in and see where the law actually stands as of January 2026.

The Difference Between Plagiarism and Copyright (Because People Mix Them Up Constantly)

Before we can talk intelligently about prompt ownership, we need to untangle two concepts that get confused almost daily in online discussions.

• Plagiarism is primarily an ethics and attribution issue. It means presenting someone else’s words or ideas as your own, even when the material could legally be used. Academic institutions, writing communities, and professional organizations care deeply about this, and rightly so.[1]

• Copyright infringement is a legal rights issue. It involves using protected expression without permission when no exception applies. Copyright focuses on original expression fixed in a tangible form and tied to human authorship.[2]

That distinction matters enormously with prompts. You can face “plagiarism concerns” in communities and forums even when copyright protection for the material is slim or nonexistent. The social consequences and the legal consequences operate on different tracks.

Can AI Prompts Be Protected by Copyright?

The short answer is: sometimes. A prompt is text, and text can be copyrighted when it reflects original human expression. At the same time, many prompts are too short or too functional to qualify.[3]

U.S. copyright protection generally requires:

• Human authorship and independent creation (you made it, you did not copy it), plus at least a minimal degree of creativity.[4]

• Original expression, rather than a mere idea, process, or method.[5]

The U.S. Copyright Office summarizes originality as “independently created by a human author” plus a “minimal degree of creativity.” The Supreme Court’s Feist Publications, Inc. v. Rural Telephone Service Co. decision from 1991 remains the classic citation for that low creativity threshold.[6]

Where Prompts Tend to Fail

A lot of prompts look something like this:

“Write a summary of this census record.”

That kind of wording is typically too thin to protect. The Copyright Office has long taken the position that names, titles, and short phrases are not protected, and this is reflected in Copyright Office regulations.[7]

Prompts also frequently function as instructions. U.S. copyright law does not extend to ideas, procedures, processes, systems, or methods of operation. So a prompt that is basically “a method for getting the model to do X” can run into the idea-expression boundary.[8]

Where Prompts Get Stronger

Prompts get closer to copyright territory when they read less like a command and more like a piece of writing, with creative choices in:

• Voice and tone (for example, a distinctive narrator or persona)
• Scene, dialogue, or a story-world setup
• Rich, original description that goes beyond what is standard or expected for the task
• A unique selection and arrangement of constraints that shows creative authorship

There is no magic word count in the law. Length matters mainly because it can be evidence that the text contains enough authorship to move beyond “short phrase” territory.[9]

A helpful analogy comes from Copyright Office guidance on recipes and directions. Circular 34 explains that recipes and formulas are generally uncopyrightable, while text directions that include enough original expression may be copyrightable. That is a useful parallel for prompts, since prompts often resemble directions for a system to follow. Circular 33 even provides an example where instructional text was refused registration because it contained insufficient creative authorship.[10]

The Book Title Question

Here is one that surprises people: Can you copyright a title like “Best AI Prompts for Genealogy 2026”?

Under U.S. practice, a single book title is not protected by copyright. The Copyright Office is very direct: copyright protection is not available for names, titles, and short phrases, even when they are distinctive.[11]

Circular 34 adds another detail that catches people off guard: even if a title appears in Copyright Office records because a larger work was registered, that record presence does not mean the title itself is copyrighted. Different works can share similar titles.[12]

What you can do instead depends on your publishing plan. The Copyright Office notes that some words or phrases can fall under trademark or unfair competition law when used to identify the source of goods or services. That is the lane many publishers explore for branding.[13]

AI-Generated Prompt Rewrites: A Legal Gray Zone

This is where things get genuinely tricky, and it is relevant to anyone using prompt optimization services like Prompt Cowboy or similar tools.

The Rewritten Prompt

In the U.S., the key issue is human authorship.

The D.C. Circuit’s Thaler v. Perlmutter decision from March 18, 2025 affirmed that the Copyright Act requires a work to be authored by a human being in the first instance. A work claimed to be autonomously generated by AI was denied registration for lack of human authorship.[14]

The Copyright Office’s AI registration guidance also states that when an AI system receives “solely a prompt” and produces complex output, the traditional elements of authorship are “determined and executed” by the technology, rather than the human user.[15]

So if a prompt-improvement service is effectively generating the rewritten prompt text with minimal human creative input, the rewritten prompt may lack the human authorship needed for copyright protection in the U.S. The analysis can be case-by-case, but the human-authorship baseline is consistent across the Copyright Office guidance and federal appellate case law.[16]

Your Original “Seed” Prompt

Your original writing can be copyrighted if it meets the normal standards: original human expression plus at least minimal creativity. If your seed prompt is very short, it may be treated as a short phrase and have little or no copyright protection.[17]

Human Editing and Curation Can Save You

Even when AI-generated material itself is excluded from protection, the Copyright Office recognizes that human-authored elements can still qualify, including:

• Creative modifications to AI-generated material
• Creative selection, coordination, or arrangement of material into a larger work.[18]

This matters for prompt businesses. Many prompt products are sold as curated libraries, workflows, or prompt packs organized by scenario, record type, jurisdiction, or research goal. That “curation layer” can be where copyright protection is more realistic, assuming the selection and arrangement reflects original authorship.[19]

A Warning About Prompt Optimization Tools

The Copyright Office’s Part 2 report from January 2025 adds a detail worth watching: it notes that in some systems, prompts may be modified or rewritten internally by the AI system, and it treats that as further distance between the human’s contribution and the final expressive result. That observation is mainly about output authorship, but it reinforces how cautious you want to be when claiming that “a tool-generated rewritten prompt” is your copyrighted text.[20]

Prompt Copyright vs. Output Copyright: Two Separate Questions

Here is an important reality check: even if your prompt text qualifies as copyrighted writing, that does not automatically make you the copyright author of the AI output.

The Copyright Office has repeatedly emphasized the human-authorship requirement for AI-generated outputs and treats prompt-only generation as lacking sufficient human control over the expressive elements of the output. The Part 2 report concludes that “given current generally available technology, prompts alone do not provide sufficient human control” to make the user the author of the output. It describes prompts as functioning like instructions that convey unprotectible ideas, and it emphasizes the gap between what a user requests and what the system actually produces.[21]

That tells us something important for prompt creators: even very detailed prompt engineering does not automatically translate into output authorship under current U.S. analysis.

Where Is This Heading?

Forecasting law is always a “watch the trend lines” exercise. Here are the trend lines supported by current official sources and recent decisions.

• The U.S. is anchored to human authorship for the foreseeable future. Between the Copyright Office’s guidance and the D.C. Circuit’s decision, the human authorship requirement is currently a stable foundation in U.S. law. Expect prompt-related strategies to work best when they emphasize clearly human-written expression, or clearly human-curated collections.[22]

• The Copyright Office is leaving room for future tech changes. The Part 2 report states that, in theory, AI systems could someday allow users to exert enough control so that the system’s contribution becomes more like rote or mechanical transcription. However, it states that the evidence does not show that level of control today. The Copyright Office also states it will continue to monitor developments and update guidance, including updates to the Compendium.[23]

• International approaches will keep diverging. Outside the U.S., there is already variation. The United Kingdom, for example, has a statute providing that for a computer-generated literary, dramatic, musical, or artistic work, the “author” is taken to be the person who undertook the arrangements necessary for creation. That legal structure differs from the U.S. approach and can affect ownership analysis for AI-assisted or computer-generated outputs.[24]

• If you sell prompt products globally, this jurisdiction spread matters, especially for contracts and enforcement.

• The biggest copyright fights around AI are still focused on training and licensing. The Copyright Office’s Part 3 report describes dozens of U.S. lawsuits and notes that legislators around the world have proposed or enacted laws regarding the use of copyrighted works in AI training. Even though prompt copyright is a narrower topic, training and licensing outcomes influence the economics of generative AI tools, which can influence how prompt libraries are built, distributed, and monetized.[25]

Practical Takeaways for Genealogists and Prompt Creators

Here are four actions that align well with the current law:

• Treat your title as branding, explore trademark instead. Plan for distinct cover design, subtitle choices, and possibly trademark strategy if this becomes a series or a long-term product line.[26]

• Build value in the human layer. Your explanations, examples, mini case studies, and genealogist-friendly guidance are more likely to qualify as protectable expression than a short command-style prompt.[27]

• Document your human authorship. If you use AI tools in drafting, keep notes showing what you wrote, what you edited, and what you selected and arranged. That approach is aligned with the Copyright Office’s registration requirements for AI-assisted works.[28]

• Assume prompt text protection is narrow. Even when a prompt is copyrighted, copyright generally targets copying of protected expression rather than broad control over an underlying method. This is a realistic expectation set by the idea-expression boundary in the statute and related Copyright Office guidance on short functional text.[29]

The Bottom Line

The legal landscape around AI prompts is still forming, but the foundations are clearer than many people realize. Human authorship remains the cornerstone of U.S. copyright law, short phrases and instructions are generally not protected, and the gap between a clever prompt and ownership of AI output is wider than prompt sellers sometimes suggest.

For genealogists building products in this space, the practical advice is straightforward: invest your creative energy in the human elements, document your authorship, and set realistic expectations about what copyright can and cannot protect.

And when someone accuses you of “stealing” their prompt? Now you will know the difference between an ethics conversation and a legal one.

Disclaimer: This article shares legal information for educational purposes. It is not legal advice. If you are planning a product launch, licensing program, or enforcement strategy, consult an IP attorney who can review your specific facts.

* * *

Author’s Note: I want to be transparent that this content – Whose Prompt Is It Anyway? Navigating AI, Copyright Law, and Prompt Plagiarism – was created in part with the help of an artificial intelligence (AI) language model – Claude Opus 4.5. The AI assisted in generating an early draft of the content, but every paragraph was subsequently reviewed, edited, and refined by me. The final content is the result of extensive human curation and creativity. I am proud to present this work and assure readers that while AI was a tool in the process, the story, style, and substance have been carefully shaped by the author.

Endnotes

[1] Ohio State University Library, “Plagiarism vs. Copyright” (https://library.osu.edu/sites/default/files/2025-04/plagiarism_vs_copyright_handout_ac_v3_2025.pdf)

[2] U.S. Copyright Office, “What Is Copyright?” (https://www.copyright.gov/what-is-copyright/)

[3] 37 C.F.R. § 202.1 (https://www.law.cornell.edu/cfr/text/37/202.1)

[4] Thaler v. Perlmutter, U.S. Court of Appeals for the D.C. Circuit, Mar. 18, 2025 (https://media.cadc.uscourts.gov/opinions/docs/2025/03/23-5233.pdf)

[5] 17 U.S.C. § 102(b) (https://www.law.cornell.edu/uscode/text/17/102)

[6] Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991) (https://tile.loc.gov/storage-services/service/ll/usrep/usrep499/usrep499340/usrep499340.pdf)

[7] U.S. Copyright Office Circular 33 (https://www.copyright.gov/circs/circ33.pdf)

[8] 17 U.S.C. § 102(b)

[9] U.S. Copyright Office Circular 33 (https://www.copyright.gov/circs/circ33.pdf); U.S. Copyright Office Circular 34 (https://www.copyright.gov/circs/circ34.pdf)

[10] Ibid.

[11] U.S. Copyright Office Circular 34 (https://www.copyright.gov/circs/circ34.pdf)

[12] Ibid.

[13] Ibid.

[14] Thaler v. Perlmutter, U.S. Court of Appeals for the D.C. Circuit, Mar. 18, 2025 (https://media.cadc.uscourts.gov/opinions/docs/2025/03/23-5233.pdf)

[15] U.S. Copyright Office, “Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence” (Mar. 2023) (https://www.copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-2-Copyrightability-Report.pdf)

[16] Ibid.

[17] 37 C.F.R. § 202.1 (https://www.law.cornell.edu/cfr/text/37/202.1)

[18] U.S. Copyright Office, “Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence” (Mar. 2023) (https://www.copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-2-Copyrightability-Report.pdf)

[19] Ibid.

[20] Ibid.

[21] Ibid.

[22] Thaler v. Perlmutter, U.S. Court of Appeals for the D.C. Circuit, Mar. 18, 2025 (https://media.cadc.uscourts.gov/opinions/docs/2025/03/23-5233.pdf)

[23] U.S. Copyright Office, “Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence” (Mar. 2023) (https://www.copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-2-Copyrightability-Report.pdf)

[24] UK Copyright, Designs and Patents Act 1988, Section 9(3) (https://www.legislation.gov.uk/ukpga/1988/48/section/9)

[25] U.S. Copyright Office, Copyright and Artificial Intelligence, Part 3: Generative AI Training (pre-publication, May 2025) (https://www.copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-3-Generative-AI-Training-Report-Pre-Publication-Version.pdf)

[26] U.S. Copyright Office Circular 34 (https://www.copyright.gov/circs/circ34.pdf)

[27] U.S. Copyright Office Circular 33 (https://www.copyright.gov/circs/circ33.pdf)

[28] U.S. Copyright Office, “Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence” (Mar. 2023) (https://www.copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-2-Copyrightability-Report.pdf)

[29] U.S. Copyright Office Circular 33 (https://www.copyright.gov/circs/circ33.pdf)