Author Created and Material Excluded Examples for AI Music Filings
Six paired examples of Author Created and Material Excluded entries for AI music workflows. Includes the Zarya of the Dawn lesson for music creators.
Freshness Check
Last reviewed Apr 3, 2026. This guide reflects the USCO's 2023 Registration Guidance and the January 2025 Part 2 report. The boundary between production utility and generative tool continues to evolve — check the Copyright Office's current guidance for your specific workflow. This guide is informational and does not constitute legal advice.
Direct Answer
When registering AI-assisted music with the U.S. Copyright Office, two fields on the eCO Standard Application determine whether your claim succeeds or fails: Author Created (what the human made) and Material Excluded (what the AI made).
Getting these fields wrong — by overclaiming human authorship or omitting AI-generated elements — can result in examiner correspondence, rejection, or a registration vulnerable to cancellation.
What Is the USCO's Two-Field Framework?
The Copyright Office's eCO Standard Application uses two complementary fields to define the scope of a copyright claim. Author Created is where you describe the copyrightable material you are claiming — it tells the Office what human-authored expression the registration will cover. Material Excluded is where you identify material in the deposit that you are not claiming, including AI-generated content.
These two fields work as a pair. Together, they draw a boundary around your claim: everything in Author Created is inside the boundary; everything in Material Excluded is outside it. The Copyright Office evaluates whether the material inside the boundary meets the human authorship and originality standards required for registration.
The 2023 Registration Guidance formalized this framework for AI-assisted works, stating that applicants must limit their claims to exclude AI-generated content. The January 2025 Part 2 report reaffirmed this approach.
Why Does Getting This Wrong Lead to Rejection?
The stakes of these two fields are higher than most applicants realize. Three failure modes are common.
Overclaiming in Author Created. If you claim "music" but the melody, harmony, and arrangement were generated by Suno, the examiner will issue correspondence asking you to narrow your claim. If you do not respond adequately, the application may be refused.
Under-excluding in Material Excluded. If AI generated the instrumental track but you leave Material Excluded blank, you have filed an inaccurate application. A registration obtained through inaccurate information may be subject to cancellation. The Office canceled portions of the Zarya of the Dawn registration for exactly this reason.
Mismatched fields. If Author Created says "lyrics and music" but Material Excluded says "AI-generated music," the examiner will flag the contradiction. Both fields must tell a coherent story.
Six Paired Examples: Author Created + Material Excluded
Each example maps a specific AI music workflow to the exact field entries you would use on the eCO Standard Application. The Author Created and Material Excluded entries are designed to work together for each scenario.
Example 1: Lyrics-Only Claim (Suno Generated Everything Else). You wrote original lyrics independently, then Suno generated melody, harmony, arrangement, and performance. Author Created: "Text / lyrics." Material Excluded: "Music (melody, harmony, arrangement); sound recording (AI-generated instrumental and vocal performance)." This registration covers the lyrics as a literary work only.
Example 2: Melody + Lyrics Claim (AI Generated Backing Track). You composed an original melody and wrote lyrics. An AI tool generated a backing track. Author Created: "Music (melody and lyrics)." Material Excluded: "Music (AI-generated harmony, bass line, rhythm); sound recording (AI-generated backing track performance)." This is a common and generally well-received filing pattern.
Example 3: Full Composition Claim with AI Production Assistance Only. You composed all musical elements and performed them yourself. AI tools were used only for mixing, mastering, or noise reduction. Author Created: "Music; lyrics; sound recording." Material Excluded: None. A Note to CO is still advisable explaining that AI tools were used only for technical audio processing. The 2023 Registration Guidance focuses on AI-generated expressive content, not AI-assisted technical processing.
Example 4: Arrangement Claim (AI Generated Raw Material, Human Arranged). You used AI to generate multiple raw musical ideas, then selected elements and arranged them into a complete composition. Author Created: "Music (arrangement, including selection and ordering of musical elements, form, dynamics, and instrumentation)." Material Excluded: "Music (AI-generated melodic fragments, chord progressions, and rhythmic patterns used as raw material)." This aligns with the Zarya of the Dawn principle that selection and arrangement can constitute copyrightable authorship.
Example 5: Sound Recording Claim (Human Performed Over AI Composition). AI generated a complete composition. You performed that composition yourself and recorded it. Author Created: "Sound recording (performance and production)." Material Excluded: "Music (AI-generated composition including melody, harmony, and arrangement)." This mirrors the practice of performers recording public domain compositions — the performance is protectable even when the underlying composition is not.
Example 6: Multiple Contributor Scenario. Two human collaborators worked with AI tools. Collaborator A wrote lyrics and melody; Collaborator B used AI to generate a harmonic arrangement and then substantially modified it. Author Created: "Music (melody by Collaborator A; modified harmonic arrangement, countermelody, and orchestration by Collaborator B); lyrics (by Collaborator A); sound recording (performance and production by both)." Material Excluded: "Music (initial AI-generated harmonic arrangement prior to human modification)."
The Zarya of the Dawn Lesson for Music Creators
In February 2023, the Copyright Office issued a decision on Zarya of the Dawn, a graphic novel that used Midjourney to generate its images. The Office registered the text and the overall selection and arrangement of text and images, but not the individual AI-generated images, because the author did not exercise sufficient control over the specific expressive output.
Three principles emerge for music creators. First, using an AI tool does not automatically disqualify your work from registration — the question is always whether identifiable human authorship exists. Second, the type of human involvement matters more than the amount of effort. Spending hours generating and selecting AI outputs does not establish authorship over the generated music; your authorship must be in the expressive content. Third, selection and arrangement can work, but the bar is meaningful — selecting two or three AI outputs and placing them in sequence is likely de minimis.
The takeaway for your filing
Be honest about what the AI generated, claim only what you can defend as human-authored expression, and provide enough detail for the examiner to evaluate your contribution.
How RightsDocket Maps Contributor Roles to These Fields
RightsDocket's claim preparation wizard collects structured information about every contributor to a work — human and AI — and maps each contribution to the appropriate Author Created and Material Excluded field entries. The system processes inputs through 56+ decision nodes to generate field-level claim language that is consistent, correctly scoped, and aligned with USCO taxonomy.
For multi-contributor scenarios, RightsDocket maps each contributor's role (composer, lyricist, arranger, performer, producer) against the specific elements they created or modified, then generates field entries that account for the full contributor table. This is the scenario where manual preparation is most error-prone — and where structured tooling provides the most value.
Frequently asked questions
What if I am not sure whether the AI's contribution needs to be excluded?
When in doubt, disclose and exclude. The 2023 Registration Guidance establishes a duty of candor regarding AI involvement. A registration with a properly scoped limitation of claim is far more defensible than one that overclaims and is later challenged or canceled.
Can I claim "music" if I wrote the melody but AI generated the harmony?
Yes, but you should be specific. Rather than claiming "music" broadly, claim "music (melody and lyrics)" and exclude "music (AI-generated harmony and arrangement)." The more precisely you scope your Author Created field, the less likely you are to draw examiner correspondence.
What is the difference between "music" and "sound recording" in these fields?
"Music" refers to the underlying musical composition — the melody, harmony, lyrics, and arrangement. "Sound recording" refers to the specific fixation of a performance — the actual audio recording. These are separate copyrightable works under U.S. copyright law. You can claim one, the other, or both, depending on your role.
Does the Copyright Office reject all AI-assisted music applications?
No. The Office has registered numerous works that involve AI assistance. The 2023 guidance and the January 2025 Part 2 report both affirm that AI-assisted works are registrable when the human contribution is identifiable and sufficient.
How long does registration take if AI is involved?
Standard eCO processing times apply. However, if your application triggers examiner correspondence, the process can extend by several months. Accurate, detailed field entries reduce the likelihood of correspondence and keep your registration on the standard timeline.
Ready To Start
Create the project record before you export.
Sign in, document contributors and AI usage, and buy a pack only when you are ready to export the filing-ready record.