Wearables Rewind Conversations Without Consent

Modern smartwatches can buffer and convert nearby speech into text or summaries (e.g., 15‑second 'Live Rewind' and continuous 'Siri Recap') by processing audio on-device and on paired phones. Those features produce stored summaries or transient transcripts that bystanders did not explicitly authorize and may run afoul of state consent laws and evolving privacy norms. — This shifts the privacy debate from always‑on cloud listening to ephemeral, device‑local audio AI, forcing legal, regulatory, and design responses about notice, consent, and auditable limits on what wearables may capture.