The voice tethering feature builds on the recent introduction of Sign Language Mode. It ensures that when an interpreter voices for a D/HH signer, captions and transcripts attribute the spoken content to the signer rather than the interpreter. This provides accurate representation in meetings and makes it clear who is contributing to the conversation. It also improves downstream meeting intelligence such as Copilot so that notes, summaries, action items, and insights are attributed to the correct participant.
GA date: March CY2026