Asked all the time

Rates are set per project and agreed in writing before you record. They reflect session length and licence scope: a short internal research licence pays differently from a broad commercial deployment. You always see the number before saying yes.

Not through Voices4AI. Cloning only happens under written consent naming the client, purpose and licence term. Nothing you record with us can be used to clone your voice unless you signed for exactly that. Our full standards are on the ethics and consent page.

They remain governed by the licence you signed: named project only, no resale, no repackaging into other datasets. Voice recordings are personal data under UK and EU law, and documented deletion routes exist.

No. Some projects need a treated space and a decent microphone, others are recorded at our UK studios, and some dataset work accepts good-quality USB setups. Tell us what you have when you apply and we match you accordingly.

Yes. Multilingual and accent dataset projects often need natural, everyday speakers rather than trained voice actors. Regional accents and under-represented languages are especially in demand - more on this in selling your voice to AI.

Send the brief through the contact page: languages, hours, speaker profiles and delivery format. Recruitment runs through the Voices4AI roster and production is delivered with Flaunt Audio.

Per-contributor consent records naming your project, the purpose and the licence term - the provenance trail procurement and EU AI Act Article 10 reviews ask for. Details on the services page.

Voices4AI is part of the 11C Media group, which also runs traditional voiceover and audio production businesses. This site exists specifically so AI work is recruited transparently, on its own terms, rather than mixed into general casting.