Anthropic prompts users to share voice data for AI training

Summary

Anthropic has begun prompting users to voluntarily share their voice conversations to enhance its AI models, stating that audio recordings and voice chat data can improve the system's understanding and response to speech. Users are informed that they can opt out or delete their data at any time through settings. The feature supports complete spoken conversations across various platforms and was recently expanded to include additional voice modes such as Opus and Sonnet, which were introduced in July 2026. This initiative aligns with Anthropic’s ongoing development of its voice capabilities, including the ability to switch between conversation models and use connected tools like Gmail and Slack when permission is granted.

Analysis