Privacy policy
What Cue captures, what it keeps, and what it never does with your meetings.
Last updated 19 July 2026
What we capture
When you run a session, Cue processes two audio streams you name up front: your voice, so it can follow your place in the script, and the shared audio of the meeting, so it can summarize what the other side says. Capture never starts silently — you confirm it first, and a persistent ● Live indicator stays visible the whole time audio is being processed.
What we keep
By default, Cue transcribes audio in flight and discards the frames — nothing is recorded. A room can opt in to retaining recordings or transcripts, in which case they are stored for a short, configurable window and can be deleted in one click. We treat the other side's data conservatively: short retention, summarized text only, and never used to train models.
Your canvas and account
Your script, canvas markup, and distilled cards live with your room so you and your partner can return to them. Host accounts are managed through magic-link or Google sign-in; guests join by link with a display name and no account. You can export a session and delete a room at any time, which removes its associated data.
Getting in touch
Have a question about your data, or want something removed? Reach us at privacy@talkoncue.com and we'll help.