Apple Watch adds AI speech transcription and conversation summarization features

Apple announced that its latest generation of Apple Watch will not retain raw audio recordings, even as it introduces capabilities that can transcribe recent speech and generate summaries of ambient conversations. The company says the new features operate by processing audio locally on the device and discarding the original sound files after analysis, a design intended to align with its broader privacy framework. The functionality, demonstrated at the recent developer conference, allows users to capture spoken notes, receive real‑time transcriptions, and obtain concise overviews of nearby dialogue without storing the underlying audio data.

Privacy experts and civil‑rights groups have raised concerns that the ability to automatically record and summarize speech, even without retaining raw audio, could create new challenges for consent and personal privacy. Critics argue that the technology may influence how individuals behave in public and private settings if they assume they could be recorded at any moment. Regulators are watching the development, and lawmakers in several jurisdictions have indicated they may examine whether existing consent laws adequately address such on‑device processing. Apple maintains that its approach balances utility with user control, but the debate over the implications of continuous, AI‑driven listening is expected to continue as the watches become widely available.