Voice Data Retention: Questions to Ask Any Dictation App

Five questions that settle it

These five questions produce specific answers, which are the kind you can hold a vendor to.

1. Does my audio leave the device, and when?
2. If audio is uploaded, how long is it kept, and is it used for training?
3. Where do transcripts live, for how long, and does deleting mean deleted?
4. What happens by default, before I change any setting?
5. Who else receives the data: model providers, analytics, human reviewers?

Ask them in email or a support ticket so you have the answer in writing. Vagueness is data too: industry-standard and may retain with no timeframe mean the answer is complicated in a direction you would not like.

Audio: uploaded, buffered, or discarded

Audio is the most sensitive part, because your voice carries who is in the room and what the background sounds like. Vendors blur three distinct cases.

• Never leaves: recognition runs on device. Verifiable by turning off the network.
• Uploaded and discarded: sent for transcription, deleted after processing. Ask for the timeframe.
• Uploaded and retained: kept for quality, training, or support. Ask how long and whether you can opt out.

Transcription does not require keeping the recording, so when a vendor keeps audio anyway, ask what you get in exchange.

Ask about failure paths too. Some apps fall back to a server when a local model is missing, mid-download, or unavailable for your language, and a local default that quietly becomes a cloud request is worth knowing about before a sensitive call.

Transcripts outlast the audio

The transcript is what contains names, salaries, medical details, and half-finished thoughts, and it usually outlives the recording by a long way. Ask:

• Is history stored on the device, in the vendor’s cloud, or both?
• Is it encrypted at rest, and can the vendor read it?
• Does deleting an entry remove it everywhere, including backups?
• Is history synced across devices, and can sync be turned off?
• Can support staff view your text when you file a ticket?

Check your own backups as well: a local history file can land in a cloud backup you set up years ago, which quietly moves data you meant to keep on the device.

Whatever app you use, find the delete path once so you are not hunting for it later. In Auri Voice it is Settings > Dictation > History, where the trash button offers Delete all history.

Good answers and red flags

Judge each answer by the default, not the option; the chooser guide makes the full case for why defaults decide what happens to most people's data.

Subprocessors, training, and human review

Data rarely stops at the vendor. AI features often call a model provider, and each hop has its own retention and access rules that a privacy page may compress into one sentence.

• Which third parties receive audio or text, and for which features?
• What are their retention periods, and are they contractually limited?
• Is any content reviewed by humans for quality or abuse handling?
• Is content used to train the vendor’s or the provider’s models?
• Where is data processed geographically, if that matters for your work?

Knowing that a small share of requests may be read by a person changes what you are willing to dictate, so ask about human review directly rather than hoping the policy volunteers it.

Ask for a subprocessor list too. Companies that serve business customers usually maintain one; if there is none, ask support to name the model providers behind the AI features.

How to verify instead of trusting the page

Most claims can be spot-checked in ten minutes, which is faster than reading a policy and shows what the software does rather than what the marketing says.

1. Download any language or engine packs, then turn off Wi-Fi.
2. Dictate two paragraphs into the app you actually write in.
3. Note what fails: recognition, insertion, or the rewrite features.
4. Look through settings for history, sync, and cloud toggles.
5. Delete one history entry and confirm it is gone after a restart.
6. Optionally run a network monitor and watch what the app contacts.

Record the results in a note with the date and app version, since behavior changes across updates.

One caution on the offline test: recognition working offline proves the recognition path is local. It does not prove nothing is uploaded when the network returns, so check sync settings too.

Where Auri stands

The same questions applied to our own products, since a guide asking you to demand specifics should offer them.

• Dictation audio: transcribed on the Mac by a local engine by default. Audio leaves only if you choose the Auri Cloud engine yourself.
• Cleanup text: sent to the provider you pick, Auri Cloud by default, and never the audio. Auri Local in AI settings keeps cleanup on the Mac (Apple Silicon).
• Dictation history: stored on the Mac. Save transcriptions is on by default and can be turned off in Settings > Dictation > History; the trash button's Delete all history clears everything. Audio history (Save audio) is off by default.
• Meeting audio: Keep for 30 days by default. The gear in Meetings > Capture and storage > Saved audio also offers Keep until I delete it and Delete after notes are ready.
• Apple Watch dictation to Mac: audio is encrypted in transit, transcribed on the Mac by a local engine, and deleted after 15 minutes.
• Auri Keyboard on iPhone: cloud dictation by default, with an On-Device option (English-only). In password and PIN fields, iOS swaps in the Apple keyboard, so no third-party keyboard sees them; many apps also block custom keyboards in payment fields, though that is each app's decision. The app's Full Access prompt says "I can't access password or credit card fields."

On training, the app's own wording is "I don't collect any data you type"; the privacy policy carries the specifics, and it is the document to hold us to. For anything beyond this list, email us the question; a written answer you can keep is worth more than a claim on a landing page, including ours.

FAQ