Why Are Songwriter Credits at the Heart of the Elvis Lawsuit Against Anthropic?

Why Are Songwriter Credits at the Heart of the Elvis Lawsuit Against Anthropic?

The Elvis Presley lawsuit against Anthropic is partly a fight about credit, not just copying. Filed October 23, 2026, the complaint alleges Anthropic stripped titles, songwriter names and publisher details from lyrics to 274 songs before training Claude. That claim, worth up to $6.85 million on its own, raises a question every publisher should ask: when AI uses your work, does your name survive?

What exactly is the credit-stripping claim?

The plaintiffs say Anthropic removed copyright management information, or CMI, from the lyrics it collected. According to Music Business Worldwide, the complaint alleges:

• Titles, songwriter credits and publisher details were removed before training
• Anthropic picked a text-extraction tool that discarded copyright notices over an alternative that kept them
• Anthropic reportedly described those notices as boilerplate

The lead plaintiff is ABG EPE IP LLC, part of Authentic Brands Group, joined by Raleigh Music Publishing, Abilene Music and trusts tied to the Aberbach family and to Luigi Creatore, co-writer of "Can't Help Falling in Love." These are allegations only. Anthropic told MBW it disagrees with the claims and intends "to defend this case fully."

Why does the law care about a songwriter's name?

US copyright law treats credits as protected information in their own right. Under 17 U.S.C. § 1202, CMI includes:

1. The title of a work and its copyright notice
2. The author's name
3. The copyright owner's name
4. Terms of use, plus numbers or links that point to this information

The statute bars intentionally removing or altering CMI when the person knows, or has reasonable grounds to know, that doing so will enable or conceal infringement. That knowledge requirement is why the complaint's detail about extraction tools matters. The plaintiffs are trying to show the credits were dropped on purpose, not lost by accident.

The money follows the statute. Section 1203 allows up to $25,000 per CMI violation, which across 274 works comes to the $6.85 million the plaintiffs plead. A separate circumvention count adds $685,000, and the main copyright claim accounts for $41.1 million.

Is this the first time credits have come up in AI music cases?

No. The October 2023 lyrics suit by Universal, Concord and ABKCO also alleged CMI removal, The Register reported. That case is ongoing. A court-approved stipulation now requires Anthropic to keep guardrails meant to stop Claude from reproducing lyrics.

The Presley complaint argues those guardrails leave a gap. It says Claude declines to print lyrics but will produce close derivative versions. That links the credit question to outputs: a rewritten song that tracks the original may arrive with nobody named as the writer.

Does fair use cover removing credits?

Fair use and credit removal are separate questions, and that's what makes the CMI count worth watching. In Bartz v. Anthropic, Judge William Alsup ruled in June 2025 that training on books was fair use but keeping pirated copies was not, according to Akin Gump. A favorable fair use ruling on training wouldn't automatically settle a claim that credits were deliberately removed.

Critics of the plaintiffs' approach have points of their own:

• The output claims rest on information and belief, and MBW notes the complaint offers no Presley-specific example
• Section 1202 requires knowledge that removal would enable or conceal infringement, a high bar to prove
• As Yahoo Finance notes, no ruling has yet resolved either training or outputs across the industry

How do AI answers handle attribution today?

AI answers don't credit sources consistently. A chat answer may quote, paraphrase or summarize material without naming where it came from. Some AI products show source links when they retrieve pages live, but text a model absorbed in training carries no built-in pointer back to its author. For a lyricist, that means a reproduced or reworked line may reach a user with no songwriter attached. For a brand, it means research, product copy or a tagline can surface without a citation that sends anyone back to the original.

The Presley case puts a legal frame around a problem that citation analysts already track. The issue isn't only whether a model uses your content. It's whether the answer tells the user where that content came from.

What does this mean for brands whose content is used without credit?

Brands should monitor attribution as closely as they monitor mentions. A model that repeats your positioning without naming you gives you no visibility, and the credit may go to whichever page it does cite.

• Measure citation, not just mention. Track whether AI answers name your brand and link to your pages when they draw on your material. MarkGrid Model Share measures how often models surface your brand compared with rivals.
• Keep credits machine-readable. Clear titles, author names and rights notices on your pages make attribution easier to retain. The MarkGrid GEO guide explains how to lay out pages so answer engines can attribute them correctly.
• Keep dated records. Screenshots of uncredited or reproduced content form the evidence base for any later claim or licensing talk.
• Watch the CMI count. If courts accept credit removal as a standalone harm, attribution becomes a legal lever as well as a visibility metric.

For the full case details, including the damages math and the wider litigation, read MarkGrid's full breakdown of the Elvis Presley lawsuit.