AI-Powered Audio Intelligence for Apple Watch Begins Testing

Jakarta (ANTARA) – A beta version of AI-powered Audio Intelligence for Apple Watch began testing on Tuesday (3 November).

According to a TechCrunch report on Wednesday (4 November), Audio Intelligence is initially being made available to developers through the iOS 27.3 operating system update.

The technology includes four main AI-powered features that allow Apple Watch Series 12 users to detect sounds, identify songs, review snippets of conversations and summarise key points from a discussion.

The sound-recognition feature is designed to help users with hearing impairments identify important sounds, such as sirens, alarms and doorbells, using their smartwatch.

The song-recognition feature uses Apple’s Shazam app to detect music playing nearby and display the song title and artist’s name.

There is also a Live Rewind feature that allows Apple Watch Series 12 and Ultra 4 users to review the previous 15 seconds of a conversation as text by double-pressing the Digital Crown.

However, the feature does not identify who is speaking, save raw audio recordings or provide a word-for-word transcript.

Apple said Siri Recaps is designed to exclude sensitive information, such as identification numbers, financial data and authentication codes, certain personal identifying information, and any language suggestive of self-harm.

Audio Intelligence on Apple Watch Series 12 and Apple Watch Ultra 4 relies on a combination of hardware, software and Apple services, including the S11 chip, built-in microphones, and Apple Intelligence and Private Cloud Compute models.