Anthropic has begun asking Claude users to voluntarily share their voice conversations to help train and improve its AI models. The request is framed as optional, but it introduces a new channel for collecting sensitive user data beyond text inputs.
Voice recordings can reveal more than words—they carry tone, accent, and other biometric-like characteristics. The announcement does not specify how long the data will be retained, who will have access to it, or what safeguards will be applied, leaving users to weigh the benefits of improved voice features against potential privacy risks.
As AI assistants become more conversational, such data collection is likely to become more common. Users who choose to participate should be aware that their spoken interactions may be used for training purposes, and those who decline may still want to review their privacy settings.