Apple’s New Watches Introduce Always-On Audio Intelligence Features, Raising Privacy Questions
On September 9, Apple announced during its keynote the addition of advanced “Audio Intelligence” features to its upcoming Apple Watch Series 12 and Ultra 4 models. These new capabilities—highlighted by the Live Rewind and Siri Recap functions—aim to elevate how users interact with day-to-day conversations, yet they immediately sparked debates regarding privacy risks and user consent for people in proximity to these devices.
How Apple’s Audio Intelligence Functions Work
The Live Rewind feature offers Watch wearers the ability to double-press the Digital Crown and receive a succinct text summary of the previous fifteen seconds of spoken discussion. In addition, Siri Recap operates by providing artificial intelligence-generated summaries of all daily transcribed conversations, delivering users an aggregated recap of what their wearable “heard” across the day.
From a privacy standpoint, Apple has outlined multiple safeguards. According to details in Apple’s audio intelligence privacy documentation, real-time deletion ensures that audio clips are never stored permanently, as local storage is continually purged. Only anonymized text is sent to Private Cloud Compute for AI processing; this text is end-to-end encrypted, and no raw recordings are retained or connected to Apple IDs after transmission.
The Series 12 and Ultra 4 devices run with the new S11 chip featuring a Secure Exclave. This enclave securely handles the initial audio analysis on the device, encrypts the content, and then sends processed data to the paired iPhone. The iPhone applies its AI model to transcribe and summarize what is received, deliberately omitting extraneous or sensitive details, and then forwards the transcript to Apple’s cloud, sharing only general information such as timestamps and approximate location. Unless the user saves it, the summary is automatically wiped after seven days. If saved, it is uploaded to iCloud and can sync with approved third-party apps such as Notes or Journal.
Challenges of Consent and Security for Bystanders
Despite robust design choices, the new audio features introduce additional concerns regarding bystander consent and security for both users and those around them. The Live Rewind option emits a short sound and briefly displays an on-screen notification to alert nearby people whenever it is triggered—although the generated text never specifies speakers by name. Conversely, Siri Recap works silently in the background, compiling conversation summaries without outwardly indicating its activity to anyone except the Watch wearer.
This silent operation may conflict with legislation in areas where every party’s express consent for audio recording is legally necessary. In the U.S., for example, twelve states stipulate that all involved individuals consent before any audio capture can take place. A number of EU member states have equivalent requirements under GDPR, prohibiting the collection of personal information without explicit notification.
Apple asserts that traditional audio recording doesn’t occur. As described in the company’s privacy statement, the Secure Exclave only retains audio temporarily before erasing it: “there is no recording to share, forward, or produce if requested by any party, because no recording exists.”
However, if a user chooses to save a summary in iCloud or sync it to a connected app, these records are then subject to the same exposure risks as any other cloud-stored data. As shown when 900,000 compromised iCloud logins came to light earlier in the year, accounts may still be susceptible to attacks, regardless of Apple’s strong infrastructure protections.
Customizing and Controlling Audio Intelligence Options
Importantly, all of these capabilities are opt-in only, and do not function by default. Users may individually activate or deactivate features to fit their personal privacy preferences or local legal mandates. To manage these tools, follow these steps:
- Sound Recognition: Open Settings > Accessibility > Sound & Name Recognition > Sound Recognition on your iPhone to toggle “Detect using Apple Watch.”
- Music Recognition: Adjust this directly from your Apple Watch at Settings > Smart Stack > Music Detection.
- Live Rewind: Use your iPhone’s Watch app and go to Siri > Live Rewind to set up Digital Crown double-press activation.
- Siri Recap: Find options in the Watch app under Siri > Siri Recap to enable daily summaries or toggle the feature as needed; for quick access, use the Apple Watch Control Center.
As Apple brings Series 12 and Ultra 4 to market, these new conversationally-aware features make the watches among the most advanced wearables yet. Still, with cloud data storage and different national regulations at play, it is essential for users to carefully consider when and how these audio features are employed—not just for their own protection, but for the privacy rights of everyone nearby.