Apple released the first developer betas of iOS 27.3 and watchOS 27.3 on Tuesday, bringing its previously announced Audio Intelligence features—including ambient, always-listening AI tools—into testing on the Apple Watch Series 12 and Apple Watch Ultra 4.
The features, first unveiled at Apple’s September “Surprise and Shine” hardware event, use the watch’s microphone and on-device AI (plus Private Cloud Compute in some cases) to process nearby speech. Two stand out for their continuous or near-continuous listening:
• Siri Recap uses ambient listening to detect conversations the wearer is part of. After a conversation ends, Apple Intelligence generates a title, short summary, and key points (not a full transcript and without speaker attribution). These appear in the Siri app. Users can schedule it (e.g., only during work hours), toggle it from Control Center, and are prompted to review and optionally save recaps, which otherwise delete after seven days. The system is designed to omit sensitive details such as financial data or government identifiers.
• Live Rewind maintains a short rolling audio buffer. A double-press of the Digital Crown displays a text snippet of roughly the previous 15 seconds of speech (useful, for example, to catch a restaurant special or directions). An audible chime and on-screen animation play even in silent mode. Unsaved snippets disappear quickly; saved ones can be kept in the Siri app or queried further with “Ask Siri.”
Both are strictly opt-in and currently limited to English. They require an Apple Watch Series 12 or Ultra 4 (powered by the S11 chip with its Secure Exclave for isolated audio handling) paired with an iPhone 16 or newer that supports Apple Intelligence and Siri AI (the iPhone 16e is excluded). They are not available to users under 13 (or, per some reports, under 18) and will not launch in the EU initially because broader
Siri AI support is not yet available there.
The features arrive amid broader industry debate over AI wearables that listen or record ambient audio (such as Meta’s smart glasses). Legal observers note potential conflicts with two-party consent eavesdropping laws in some U.S. states, since bystanders may not receive clear notice that speech is being processed. Apple’s approach relies on opt-in controls, short buffers, no persistent raw recordings, and visible/audible cues for Live Rewind rather than continuous recording.
MacDailyNews Take: We love these features, especially their privacy protections. Raw audio is processed inside a dedicated hardware-protected buffer on the Apple Watch’s S11 chip (the Secure Exclave), is not saved as a file, and is not accessible to the operating system, apps, or Apple. Condensed text for Siri Recap is encrypted and sent to Private Cloud Compute for summarization, after which the intermediate data is deleted. Saved content syncs end-to-end encrypted via iCloud when two-factor authentication and a device passcode are enabled. Apple also published a technical privacy overview and a support article for users here: About Audio Intelligence features on Apple Watch Series 12 and Apple Watch Ultra 4.
Developer betas are available now. A public beta is expected to follow, with a full consumer release targeted for December. Battery impact is described as negligible. The features remain in testing, so behavior, language support, and availability may still change before the public launch.
Please help support MacDailyNews — and enjoy subscriber-only articles, comments, chat, and more — by subscribing to our Substack: macdailynews.substack.com. Thank you!
Support MacDailyNews at no extra cost to you by using this link to shop at Amazon.
