Tech

Apple Watch Series 12 and Ultra 4 get controversial AI features in watchOS

Apple is rolling out watchOS 27.3 developer beta with controversial AI features Live Rewind and Siri Recap for Series 12 and Ultra 4.

·2 min read
Apple Watch Series 12 and Ultra 4 get controversial AI features in watchOS

Apple has begun rolling out the much-discussed microphone-based AI features for its latest smartwatches. Live Rewind and Siri Recap are now arriving as part of the first watchOS 27.3 developer beta, expanding the Audio Intelligence suite originally introduced with the Apple Watch Series 12 and Watch Ultra 4.

  • Live Rewind displays a text snippet of the last 15 seconds of audio upon double-pressing the Digital Crown.
  • Siri Recap ambiently takes notes on daily conversations and generates summaries in the Siri app.
  • Data is protected using hardware-isolated Secure Exclave and Private Cloud Compute for strict privacy.
  • Features currently require English language support and an iPhone 16 or newer paired device.

How Live Rewind works on Apple Watch

The Live Rewind feature is designed to capture speech you might have missed by showing the last 15 seconds of audio as text on the watch face. Users can trigger this by double-pressing the Digital Crown when someone is listing details or giving directions. The resulting snippet can be queried via Siri or saved directly.

For privacy and transparency, using Live Rewind triggers an audible chime—even in silent mode—alongside a full-screen animation and a microphone indicator so nearby individuals know it is active. If ignored, the text disappears shortly after the display dims.

Read alsowatchOS 27.3 beta adds new Siri Recap and Live Rewind features for Apple Watch

Siri Recap and privacy safeguards

Siri Recap takes notes on conversations throughout the day, generating titles, summaries, and key points in the Recaps tab. It avoids creating verbatim transcripts and is intentionally programmed to omit sensitive information like financial data. Recaps automatically delete after seven days unless manually saved.

Apple emphasizes that no raw audio recordings are ever created. Unprocessed audio enters a hardware-isolated Secure Exclave on the S11 chip and paired iPhone, preventing access by the OS or Apple itself. For summaries, a condensed transcript is processed via Private Cloud Compute.

Availability and requirements

Both features are strictly opt-in and require an Apple Watch Series 12 or Ultra 4 paired with an iPhone 16 or newer running Apple Intelligence. They are currently restricted to English, unavailable for users under 13, and initially excluded from the EU due to local AI regulations and availability.

Source: GSMArena.com - Latest articles

Related articles