// ZDNET — MOBILE & WEB
Your Apple Watch has 2 new controversial listening features – but you have to opt in
Siri Recap and Live Rewind are finally here, two months after Apple unveiled the always-listening Apple Watch features. Here's how they work.
Siri Recap and Live Rewind, Apple’s two controversial, always-listening Audio Intelligence features, are now available on the Apple Watch Series 12 and Apple Watch Ultra 4 with the iOS 27.3 developer beta. Both features are opt-in.
Apple unveiled Siri Recap and Live Rewind in September alongside its Series 12 and Ultra 4 smartwatches, but it waited to release the features until after the smartwatches launched. Siri Recap provides a high-level summary of meetings, while Live Rewind continuously buffers audio so the watch wearer can pull up a transcript of the last 15 seconds.
Also: Apple’s Audio Intelligence is priming us for its real AI wearable future
To try both features today, update your iPhone to the iOS 27.3 developer beta, and then update your Apple Watch.
Within the Siri app, users will find a new section called “Recaps,” where both the Siri Recaps and Live Rewinds live.
Also: Apple Watch 12 can listen in and recap your conversations now – genius or privacy nightmare?
Siri Recap isn’t listening for nitty-gritty details within a meeting or recording a transcript’s worth of text. It’s more geared toward a conversation’s key takeaways, like important dates, times, prices, and action items. A demo of the feature showed an off-the-cuff conversation between a parent and a child discussing college application strategies. Siri Recap took notes on the high school activities the student could pursue to include in their application, the topics they could write about in their essay, and the colleges they could consider applying to.
After a conversation, a user can save the meeting notes in the Siri app or export them to Notes or Journal. If they don’t save the conversation, it’s automatically deleted after seven days.
Apple designed the feature to exclude certain information from the Siri Recap. It includes a person’s name only if they introduce themselves or if their name is regularly brought up in conversation. Otherwise, it avoids identifying speakers. It also scrubs the audio of language that promotes self-harm, financial and authentication data, and personal identifiers.