Read the Apple document explaining how new listening features still protect your privacy

Details

Apple used its iPhone Duo launch event on September 9, 2026 to debut several ambient-listening Siri features: Siri Recap, Live Rewind, Sound Recognition, and Music Recognition. Anticipating privacy concerns, Apple simultaneously released a technical document detailing how audio from these features is handled. The core protection relies on the Secure Exclave — a hardware-isolated compartment in the new S11 chip powering Apple Watch Series 12 and Apple Watch Ultra 4. Raw audio enters the Secure Exclave, is processed for speech, sounds, or music, and is continuously overwritten in a buffer that never creates an audio file. The data cannot be accessed by watchOS, apps, the user, or Apple itself. Several user-controlled safeguards supplement the hardware isolation. Live Rewind requires an explicit double-tap of the Digital Crown for each session, preventing passive background recording. If processed data moves from Apple Watch to iPhone, it travels with end-to-end encryption through both devices' Secure Exclaves. Text saved from Siri Recap or Live Rewind syncs with end-to-end encryption, provided users have a device passcode and iCloud two-factor authentication enabled. Apple's document stops short of addressing third-party audits or clarifying retention policies for derived transcripts — as opposed to raw audio — which privacy researchers are likely to flag as remaining gaps.