Anthropic has started asking Claude users to voluntarily share their voice conversations to help train and improve its artificial intelligence models. The new prompt appears when users access Claude voice features. It explicitly asks users to allow Anthropic to use their voice data for artificial intelligence training.
Audio recordings and voice chat data help improve how Anthropic artificial intelligence models understand and respond to speech. The prompt informs users that they can turn this off or delete this data anytime in settings. This data collection appears to be entirely optional for users.
Voice settings operate independently
Users can select Allow or Not now when the prompt appears. A dedicated toggle sits inside the Claude Privacy settings menu. Voice training is separate from Anthropic existing option to use regular chats and coding sessions for model training.
Users can allow Anthropic to train on voice conversations while keeping regular chats and Claude Code sessions excluded. The reverse choice is also possible. Inside Settings and Privacy, Anthropic describes the option as Allow us to use your voice data.
Read nextAnthropic launches Claude Frontier Academy with $100 million training commitmentFor approved sessions, the setting allows the use of audio recordings and voice chat data to improve Anthropic artificial intelligence models. The setting is disabled by default in the tested interface. Anthropic does not automatically opt users into sharing their voice recordings.
Users can turn the option off later. Anthropic states that users can delete the voice data from settings if they change their mind.
