Get a license or do not sample, a court held in 2005

A sampling ruling and the Copyright Office’s separate work on digital replicas show why a cleared sample and a generated sound-alike differ.

First source published

June 3, 2005

Site publication

September 18, 2026

What happened

On 3 June 2005, the Sixth Circuit decided Bridgeport Music, Inc. v. Dimension Films, holding that a sound recording copyright owner’s exclusive right to sample their own recording is absolute: “Get a license or do not sample,” the court wrote, rejecting a de minimis defence for a two-second guitar sample. Separately, the U.S. Copyright Office has been developing guidance on digital replicas – technology that imitates a real person’s voice or likeness – publishing Part 1 of its AI report on that subject on 31 July 2024, distinct from its later work on generated-output copyrightability.

What the documents say

The Bridgeport opinion turns on 17 U.S.C. § 114(b), read as giving a sound recording’s owner rights over “the actual sounds fixed in the recording,” but not extending to “another sound recording that consists entirely of an independent fixation of other sounds, even though such sounds imitate or simulate those in the copyrighted sound recording.” That sentence is the whole distinction: copying the actual captured audio needs a licence; an independently created recording that merely sounds similar does not infringe the sound recording copyright, however close the imitation. The Music Modernization Act summary reinforces the separateness of a musical work’s copyright from a sound recording’s, since a resemblance in melody is a different question, governed by a different right, from a resemblance in captured sound.

Why it matters for makers

The mechanism a producer needs is exactly this fork: sampling takes the actual fixed sound and needs a licence, full stop, under Bridgeport’s reading. A generated sound-alike – a model producing new audio that resembles a voice or production style without reproducing any specific recorded sample – is not sampling in that sense, because no sound recording was copied. But not sampling is not the same as no rights implicated: the underlying composition can still be infringed if substantially copied, and a real person’s voice can separately be protected by publicity or digital-replica-style rights that Bridgeport’s analysis never addressed, as the Office’s separate digital-replicas work makes clear.

What to check before you use it

This is an editorial reading of the cited documents, not legal advice, and not a claim that any named artist has been infringed. If actual recorded audio was copied or rearranged, treat Bridgeport’s bright-line rule as controlling: clear it or do not use it. If the output is an independent, model-generated approximation, separately check whether the composition was substantially reproduced, and whether the voice or persona imitated is a real, identifiable individual protected under a publicity or digital-replica-style right in their jurisdiction.

• Was any actual recorded sound reproduced or rearranged, or was the audio independently generated?
• Could the underlying melody or composition, separate from the recording, be substantially similar?
• Is a real, identifiable person’s voice or persona being imitated closely enough to raise a separate right?

Bridgeport settled one narrow question about captured sound two decades ago; it did not anticipate a tool that generates new audio without copying any recording. Reading the two document types side by side is the way to see which of several distinct rights, if any, is in play.

Sources & reading trail

Holds that any unlicensed sampling of a sound recording, however small, infringes, while clarifying that an independently created recording that merely imitates does not.

Source published: 3 June 2005 · Retrieved: 16 September 2026

Describes the separate legal treatment of musical works and sound recordings, including the January 2021 start of mandatory Mechanical Licensing Collective registration.

Source published: Not established · Retrieved: 16 September 2026

States the Office's Part 1 report on digital replicas, addressing imitation of a person's voice or likeness, was published 31 July 2024, separate from its sound-recording sampling doctrine.

Source published: 31 July 2024 · Retrieved: 16 September 2026

Papers, terms and official documents establish the record; the maker reading and the checks are Signal to Song editorial analysis. This retrospective draft does not imply the site published on the event date.

Continue reading

• YouTube's own blog places Content ID's launch in 2007
• A blanket licence created a single place to register a song
• The Copyright Office ruled prompts alone do not make an author
• Browse the complete the archive

Published September 18, 2026, not on the date of the event described.