To turn speech into text, we process your recording. Here’s a factual overview of how that works.
- On Apple-silicon Macs your dictation audio stays on your device. Where the cloud does the work, audio and content run through servers in Germany / the EU.
- Audio is never stored. We keep your text for 24 hours, permanently or not at all — you decide; organisations set it centrally.
- We never train AI on your data, and neither does our AI vendor.
- Your custom dictionary stays linked to your account so terms work across devices.
