Apple unveiled its new hardware lineup at its “Surprise and Shine” event, but the most contentious development was not a device design. It was the Apple Watch’s expanding ability to process nearby sound, transcribe snippets of speech and produce AI-generated conversation notes.¹

The company’s least controversial offering is Audio Intelligence, an on-device feature intended to alert wearers to practical sounds including alarms, sirens, doorbells and a crying baby. For deaf or hard-of-hearing users, that could be a meaningful accessibility tool — and, in an emergency, potentially lifesaving.¹

The tension sharpens with Live Rewind. By double-pressing the Digital Crown, users can go back 15 seconds and transcribe what was just said, then save the text in Apple’s standalone Siri app. Apple pitches the tool for retrieving a book recommendation or a colleague’s idea that a wearer missed the first time. But critics argue it also makes it easy to preserve another person’s unguarded remark without the conspicuous act of raising a phone to record.¹

Siri Recap goes further, using ambient listening to create “high-level notes” with titles, summaries and key points. Apple says the feature is optional rather than permanently active, can be limited to certain times or places, and can be toggled off from Control Center. It also says it neither creates nor stores raw audio, does not identify speakers, and protects recaps and transcripts with end-to-end encryption.¹

Those safeguards mark Apple’s attempt to distinguish its approach from always-on AI wearables. Yet the central criticism is cultural as much as technical: a watch gesture may be too unobtrusive for bystanders to recognize when their words are being turned into a record. The result, critics warn, could be a privacy-conscious company normalizing the idea that technology is always listening.¹