Privacy Policy
Summary
BlurPOV is built so your video never has to leave your device. We don't run a server, we don't have an account system, and we don't build a profile of you for our own purposes. Detection, identification, and rendering all happen locally on your iPhone.
To keep the app free, BlurPOV shows ads served by Google AdMob. AdMob is a third-party SDK that collects standard advertising-industry data (described in the "Advertising and ads SDK" section below). It does not have access to your videos, audio, transcripts, or face data — those still never leave your device. This page explains both halves in detail.
What we process
BlurPOV processes the following categories of data only on your device, only for as long as needed to deliver the feature you asked for:
- Camera frames. When you use live mode, the camera feed is analyzed in memory to detect and blur faces. Frames are not recorded unless you tell the app to capture them.
- Videos you import. When you pick a clip from Photos or Files, BlurPOV reads the video, scans it for faces, and writes a redacted output back to a file you choose to keep.
- Face crops and embeddings. While scanning, BlurPOV creates small thumbnail images and numeric face embeddings to cluster appearances of the same person. These exist in memory and in temporary files for the duration of your session.
- Manual blur regions. If you record drag-while-playing blurs, the rectangles and timestamps you draw are stored alongside your in-progress edit.
- Audio from videos you scan. If you choose to use the audio profanity filter, BlurPOV extracts the audio track of the clip you selected to a temporary file and transcribes it on your device to find user-defined words you have asked the app to bleep.
- Microphone audio (live mode only). When you record a new clip with the in-app camera, the microphone is captured alongside the video and written into the recording you save. It is not streamed anywhere.
- Your custom word list. The list of words you have asked BlurPOV to flag and bleep is stored locally in the app's settings on your device.
Face data
Because BlurPOV's purpose is to detect and blur faces, this section spells out exactly how the app handles face data.
- What face data is collected. While you are scanning a clip or using live camera mode, BlurPOV generates face bounding boxes, small face thumbnail crops, and numeric face embeddings (feature vectors used to tell appearances of the same person apart). It does not collect names, identities, demographic information, or any biometric template intended to identify a specific real-world person.
- How face data is used. Face data is used solely to (1) detect faces so they can be blurred, (2) cluster appearances of the same person across a video so you can make one redact-or-reveal decision per individual, and (3) render a blurred output video. It is never used for advertising, analytics, profiling, model training, or identification against any external database.
- Sharing and disclosure of face data. Face data is not shared with any third party. It is not transmitted off your device. The app contains no networking code that uploads face data, and no third-party analytics, advertising, or identity-matching SDK has access to it.
- Where face data is stored. Face data exists only on your iPhone — in memory while you are editing, and in your device's temporary directory for intermediate files created while scanning or rendering. There is no server, no account, and no remote storage of face data.
- How long face data is retained. Face data is retained only for the duration of your editing session. Bounding boxes, thumbnails, and embeddings are released from memory when you finish or dismiss the edit or close the app. Temporary files are removed by iOS or when you close the project or reinstall the app. The only artifact that persists is the rendered output video you choose to save to your Photos library, which contains blurred faces — not raw face data.
Audio data
BlurPOV includes optional features that work with audio: an audio profanity filter that bleeps spoken words you have flagged, manual beep regions you can draw on a clip's timeline, and the microphone capture used by the in-app camera. This section spells out exactly how the app handles audio.
- What audio data is collected. When you run the audio profanity filter on a clip, BlurPOV extracts that clip's audio track to a temporary file and runs speech recognition on it to produce a transcript with timestamps. When you record with the in-app camera, microphone audio is captured into the resulting video file. When you draw a manual beep region, the timestamps of that region are stored with your edit. BlurPOV does not collect voiceprints or any biometric audio template intended to identify a specific real-world speaker.
- How audio data is used. Audio data is used solely to (1) transcribe speech so the app can locate words you have added to your custom filter list, (2) generate a beeped output where those words and any manual beep regions you drew are replaced with a tone, and (3) preserve audio in clips you record with the in-app camera. It is never used for advertising, analytics, profiling, model training, voice identification, or matching against any external database.
- On-device speech recognition. BlurPOV requests on-device speech recognition from iOS so that the audio of the clip you scan does not leave your iPhone during transcription. If on-device recognition is not supported for your language on your device, BlurPOV will surface an error rather than fall back to cloud recognition in shipping builds.
- Sharing and disclosure of audio data. Audio data is not shared with any third party. It is not transmitted off your device. The app contains no networking code that uploads audio, transcripts, or word lists, and no third-party analytics, advertising, or voice-recognition SDK has access to them.
- Where audio data is stored. Extracted audio for transcription, generated beep tones, and intermediate render files exist only on your iPhone — in memory while you are editing and in your device's temporary directory while a scan or render is in progress. Your custom word list is stored in the app's local settings on your device. There is no server, no account, and no remote storage of audio, transcripts, or word lists.
- How long audio data is retained. The temporary audio file extracted for transcription is deleted as soon as transcription finishes. The transcript itself exists only in memory during the scan and is discarded once the flagged ranges have been computed. Beep-tone files and intermediate render files are removed when the export completes. The only audio artifact that persists is the bleeped output video you choose to save to your Photos library, which contains your original audio with flagged ranges replaced by a beep tone — not a stored transcript. Your custom word list persists until you remove the words or delete the app.
- Custom word lists. BlurPOV ships with no preset profanity list. Every word that gets flagged is one you added yourself. You can disable, re-enable, or delete words at any time from the app's profanity settings.
Advertising and ads SDK
BlurPOV displays ads served by Google AdMob to keep the app free. AdMob is a third-party SDK from Google LLC. Unlike the on-device processing described above, AdMob does communicate with Google's servers in order to fetch and measure ads.
- What AdMob may collect. A device-scoped advertising identifier (the iOS Advertising Identifier, or IDFA) — but only if you grant App Tracking Transparency permission when iOS prompts you. If you decline, AdMob does not receive your IDFA. Either way, AdMob also collects coarse device characteristics such as your iPhone model, iOS version, and language; an approximate location derived from your IP address (not GPS); ad-interaction events such as which ads were shown and tapped; and a randomly generated session identifier.
- What AdMob is used for. Selecting which ads to show you, measuring whether ads are seen and tapped, attribution (so the advertiser knows you came from BlurPOV), and frequency capping so you don't see the same ad over and over. AdMob does not have access to your videos, audio, transcripts, face data, embeddings, or any of the on-device editing artifacts described in the sections above.
- Your tracking choice. When you first launch BlurPOV, iOS will ask whether you allow tracking. You can decline. Ads still appear if you decline, but they are non-personalized and AdMob does not receive your IDFA. You can change this choice at any time in iOS Settings → Privacy & Security → Tracking → BlurPOV.
- SKAdNetwork. BlurPOV also participates in Apple's SKAdNetwork, an iOS-level framework that lets advertisers measure the effectiveness of an ad campaign without identifying you. SKAdNetwork is privacy-preserving by design.
- Google's privacy policy. For full details on how Google handles the data AdMob collects, see Google's privacy policy at policies.google.com/privacy and AdMob's specific disclosures at support.google.com/admob/answer/6128543.
What we don't do
- We do not upload your videos, frames, faces, embeddings, audio, transcripts, or word lists to any server.
- We do not maintain a face database, voiceprint database, or attempt to match your faces or voices against any external identity service.
- We do not require an account, email address, phone number, or login.
- We do not sell or rent personal information. AdMob may use the data it collects for its own ad targeting and measurement, as described in "Advertising and ads SDK" above.
- We do not include third-party analytics SDKs. The only third-party SDK in BlurPOV is Google AdMob, used solely to serve the ads described above.
Permissions BlurPOV asks for
- Camera. Required for live redaction mode. The feed is processed locally and is not transmitted.
- Microphone. Required to capture sound when you record a clip with the in-app camera. The recording is written to your device and is not transmitted.
- Speech recognition. Required by the optional audio profanity filter. BlurPOV asks iOS for on-device speech recognition so the audio of the clip you scan stays on your iPhone during transcription.
- Photo library (read). Required to import a video you select for redaction. Only the asset you pick is read.
- Photo library (add). Required to save a rendered, redacted video back to your library.
- App Tracking Transparency. Optional. When you first open BlurPOV, iOS will ask whether you allow tracking. Granting permission lets AdMob deliver more relevant ads via your iOS Advertising Identifier. Declining does not disable any feature of the app — only ad relevance.
You can change these permissions at any time in iOS Settings → BlurPOV. Denying a permission disables the feature that uses it but does not affect anything else in the app.
Storage and retention
Source clips you import and intermediate files BlurPOV creates while scanning or rendering are written to your device's temporary directory. iOS may delete this directory on its own; you can also clear it by closing the app or reinstalling. Output videos saved to your Photos library remain there until you delete them — they're yours.
Children
BlurPOV is intended for general audiences and is not directed at children under 13 (or under 16 in the EEA and UK). We do not knowingly collect personal information from children, and the ads served in BlurPOV are configured for general audiences only. If you are a parent or guardian using BlurPOV to redact footage that includes a minor, the redacted output stays under your control on your device.
Security
Because data does not leave your device, the practical security of your BlurPOV content depends on the security of your device — your passcode, Face ID or Touch ID, iCloud settings, and how you choose to share any rendered videos. We recommend keeping iOS up to date.
Crash reports and diagnostics
If you have opted in to share diagnostics with Apple in iOS Settings, Apple may collect anonymized crash logs that include the BlurPOV process. Those logs are governed by Apple's privacy policy, not ours, and they do not contain your video content.
Changes to this policy
If we change how BlurPOV handles data — for example, by introducing an optional cloud feature — we will update this policy and change the effective date at the top. Material changes will be surfaced inside the app before they take effect. Continuing to use BlurPOV after a change means you accept the updated policy.
Contact
Questions about this policy or how BlurPOV handles data? Email support@blurpov.app.