ElevenLabs
ElevenLabs says certain submitted data may be used to improve audio models unless the account disables model improvement; Enterprise customer data is not used for that purpose by default.
Submitted audio, text, video, preferences, feedback, and related metadata may be processed to research, develop, train, or improve models when the improvement setting is on. Enterprise data is excluded from that training path by default.
Open account Terms and privacy settings and disable 'Improve the models for everyone.' The change is prospective and does not undo prior training uses.
ElevenLabs says it does not use provided data to clone a voice without consent or create advertising profiles, and describes training practices designed to disassociate or remove identifying data.
The cited training disclosure does not provide one universal content-retention period. Opt-out affects future submitted data rather than prior uses.
The reviewed training disclosure does not provide one general human-review statement.
Foundational reference pages for teams adopting, deploying, and operating production AI with inspectable controls.