Privacy Policy

Information We DO NOT Collect

iTranscription does NOT collect:

• Personal identification information (name, email, phone number)
• Audio recordings
• Transcription text or content
• Usage data or analytics
• Device identifiers or advertising IDs
• Location data
• Contact lists or address books
• Photos or media library access (except files you explicitly select)
• Crash reports or diagnostic data
• Any other personally identifiable information

Local Data Processing

All audio processing, transcription, summarization, and translation happens entirely on your device:

• Recording: Audio recordings are stored only in the app's sandboxed storage on your device
• Transcription: OpenAI's Whisper AI models run locally on your device to convert speech to text
• Summarization: AI summarization (iOS 18+) uses on-device language models via the Cactus framework
• Translation: AI translation (iOS 18+) uses on-device language models via the Cactus framework
• No Internet Required: Once AI models are downloaded, the app functions completely offline

Data Storage

All data is stored locally in your device's secure app sandbox:

• Audio Files: Recordings and imported files are stored in the app's private documents directory
• Transcriptions: Transcription text and metadata are stored in a local SQLite database
• Settings: User preferences are stored locally using iOS secure storage mechanisms
• AI Models: Downloaded models are stored in the app's cache directory

Your data is never uploaded to external servers, cloud storage, or transmitted over the internet.

Network Usage

The app uses your internet connection only for:

• Downloading AI Models: When you first use the app or download additional models, the app retrieves them from Hugging Face (huggingface.co)
• Model Updates: Optional model updates are fetched from Hugging Face when available

What is NOT transmitted over the network:

• Your audio recordings
• Your transcription text
• Your personal information
• Any usage or analytics data

Permissions Required

Microphone Access:

• Purpose: To record audio when you use the recording feature
• Control: You can deny or revoke this permission at any time in iOS Settings
• Usage: Audio is recorded locally and never transmitted externally

File System Access:

• Purpose: To import audio/video files you select for transcription
• Control: You choose which files to import; the app cannot access files without your explicit selection
• Usage: Imported files are copied to the app's private storage

Local Notifications (Optional):

• Purpose: To notify you when transcription jobs complete
• Control: You can disable notifications at any time in iOS Settings
• Usage: All notifications are generated locally; no external notification service is used

Data Retention and Deletion

• User Control: You have complete control over all data in the app
• Deletion: You can delete any recording, transcription, or imported file at any time

Third-Party Services

The only third-party service we interact with is:

Hugging Face (huggingface.co)

• Purpose: Hosting and distribution of open-source AI models (Whisper and language models)
• Data Shared: Only anonymous download requests for model files
• Privacy Policy: https://huggingface.co/privacy

We do not use any third-party analytics, advertising, tracking, or cloud services.