Copyright: turning a text into audio is reproducing it

It is tempting to think of the audio version as a bit of comfort bolted onto an article that is already online. In copyright terms it is a more precise act: having a voice read a text and recording the result reproduces that text in another form. This is not a distant legal footnote, it is the question that decides who is allowed to make the audio at all. This piece asks it plainly, cites its sources, and sets out what WeDispatch asks of a customer at the moment they create an audio. As a French service, we point to French law here; the principle, that turning text into audio is an act of reproduction, is familiar across copyright systems, but the specifics are your own situation's.

Why reading a text aloud is a reproduction

The French Intellectual Property Code defines an author's exploitation right as the sum of two rights: the right of representation and the right of reproduction (article L. 122-1). Reproduction, article L. 122-3 spells out, is the material fixation of the work by any process that lets it be communicated to the public indirectly. An audio file that fixes an article read aloud falls squarely inside that definition: it is a fixation, on a new medium, of the written work.

The consequence is set by article L. 122-4: any reproduction made without the author's consent (or their rightholders') is unlawful. In other words, producing the audio version of a text requires holding the rights to it, or being authorised. For your own article, the question does not arise. It arises the moment you turn into audio a text you did not write.

Three concrete cases, three answers

The first case is the simplest: your own article, written by your newsroom. You hold the rights, you can have it read and distribute the audio, as a podcast included. Nothing to ask anyone.

The second case is the agency wire story. A subscription to a press agency grants a use governed by a contract, which rarely provides on its own for voicing and re-distribution in audio form. Before voicing a wire feed, check what the contract allows: the right to publish the text does not automatically carry the right to produce and distribute an audio version of it.

The third case is a guest op-ed, or a freelancer's piece. The author remains the holder of their rights absent a written assignment. Publishing the op-ed on the site is one agreement; making it a distributed audio file is a further use, better provided for in advance. The right instinct is to fold voicing into the original agreement rather than treat it as a given.

What the customer declares to WeDispatch

We are not in the publisher's place to judge their rights, and we do not pretend to be. Our terms of sale set out a clear declaration: before a first creation, and when connecting a site, the customer declares that they hold the rights to the content they have read and published, or that they have authorisation from its authors and publishers. That declaration covers the voicing and distribution of the audio files, podcast included. The customer warrants that they hold all the necessary rights and that the content is lawful.

In return, our position on ownership is clear: the customer keeps the entirety of the rights to their content and to the generated audio files. We claim nothing over what we help produce. And on a substantiated report from a rightholder, we can remove content and its audio files without delay, and suspend the account in the event of repeated breaches. That is not a threat, it is the logical counterpart of the declaration: it would mean nothing if a well-founded report had no effect.

The private document, never published

There is one use where the question of distribution does not arise: a document added for private listening. A report, an internal memo, a PDF you want to listen to for yourself and not publish. That document never enters a podcast feed, a public page or an integration: it stays inside the account's personal listening. French law knows this distinction as the private-copy exception (article L. 122-5, 2° of the Intellectual Property Code), which covers reproductions strictly reserved for the copier's private use. The exception has its limits and does not cover redistribution; but it makes clear why listening to a document yourself and publishing it for everyone are not the same act, neither in use nor in rights.

What to take away

Turning a text into audio is reproducing it, and reproduction requires the rights. For your own articles, you have them. For a wire story or an op-ed, check first, because the right to publish the text does not carry the right to distribute its audio. WeDispatch asks you to declare that you hold those rights, leaves you their full ownership, and acts on a rightholder's report. For what happens between your text and what gets said, and where it travels during production, our page on French text to speech sets out the approach. This piece gives you bearings, it does not replace a lawyer's view of your own situation.

Read next

• Listening to an article with the screen locked, and in the car
• A synthetic voice: should you disclose it? What the EU rules say
• One podcast per section, each configurable, on Apple Podcasts and Spotify