Data Handling (Infinity Transcribe Recorder)

⚠️ Infinity Transcribe Recorder uses services from Google (Gemini) and OpenAI (Speech to Text API). How data is handled largely depends on the terms of use of these two companies.

Introduction

Summary

• Infinity Transcribe Recorder does not use any data for AI training.
• Data retention is as follows:
  • Due to Google's specifications, data is retained on Google's side for 55 days for the purpose of abuse monitoring.
  • Our service’s servers (including logs) generally do not store audio or text data.
  • However, when using the backup feature, data is stored on Firebase Storage.

Details

1. Handling when transcribing audio data

Audio data is processed using Google’s Gemini or OpenAI’s Speech to Text API (hereafter, STT API). We primarily use the Gemini API, but if repeated processing errors occur, we fall back to the OpenAI STT API.

Data handling via each company’s API is as follows.

• Google Gemini API (used as a paid service for Infinity Transcribe Recorder) does not use your data for AI training. Data is retained for 55 days for the purpose of abuse monitoring.

  When using the paid service (including paid tiers of the Gemini API), Google does not use your prompts (including related system instructions, cached content, and files such as images, videos, and documents) or responses to improve their products.

  • Terms: https://ai.google.dev/gemini-api/terms
  • To assist with monitoring for abuse, Google retains the following data for 55 days:
    • Prompts: Text prompts you send to the API
    • Context information: Additional context provided along with the prompt
    • Output: Responses generated by the Gemini API
    • Documentation: https://ai.google.dev/gemini-api/docs/abuse-monitoring

• For the OpenAI STT API (/v1/audio/transcriptions), your data is not used for AI training and is not retained for abuse monitoring.

  Your data is your data. As of March 1, 2023, data sent to the OpenAI API is not used to train or improve OpenAI models (unless you explicitly opt in to share data, e.g., by providing feedback in Playground). One benefit of opting in is that models may improve for your use case over time. To detect misuse, API data may be retained for up to 30 days and then deleted (unless otherwise required by law). For trusted customers with highly sensitive applications, zero data retention may be available. With zero data retention, request and response bodies are not stored in any logging mechanism and exist only in memory to serve the request. Note that this data policy does not apply to OpenAI’s non-API consumer services such as ChatGPT and DALL·E Labs. Default usage policies by endpoint (excerpt) (/v1/audio/transcriptions) Training: No; Default retention: Zero; Eligibility for zero retention: — (excerpt).

  • Data policy: https://platform.openai.com/docs/models/how-we-use-your-data#how-we-use-your-data

Our service’s servers do not store the audio or text data used by the service. These user-provided data are also not stored in usage logs.

2. Handling during backups

Infinity Transcribe Recorder provides a backup feature. This uses a cloud service called Firebase Storage.

Data is logically segmented and managed using individual identifiers issued by Auth. In addition, Firebase Security Rules prevent viewing when the user ID does not match, so it is fundamentally impossible for other users to access your data.

If an investigation related to the backup feature becomes necessary, the Infinity Transcribe Recorder operations team may view data after confirming the user’s consent. The app is usable without the backup feature; if you have concerns, we recommend using the app without enabling backups.

3. User behavioral data

Infinity Transcribe Recorder collects user behavioral data using Firebase Analytics for the purposes of service improvement, abuse prevention, and stable app operations.

• Data collected
  • App launch count and timestamps
  • Usage events such as screen transitions and button operations
  • App crash reports and error information
  • Device information (OS version, model, language settings, etc.)

These behavioral data are collected in anonymized and statistical form, not linked to personally identifiable information (such as name, address, or email). They are used primarily to enhance functionality and quality, and to investigate or prevent incidents.

• Collection and transmission

  • Tool used: Firebase Analytics (an analytics service provided by Google)
  • Destination: Google’s servers
  • Information is transmitted in an anonymized form, such as the device’s unique identifier.
  • The transmitted data is reviewed and analyzed by the provider via reporting features such as usage reports and crash analysis.

• Purposes of use

  • Improve app usability and performance
  • Understand service usage
  • Prevent and detect abuse and incidents
  • Inform new feature development and improve usability

Terms

Based on the above, these points are reflected in the Terms of Use and Privacy Policy.

• Terms of Use
• Privacy Policy

That’s all.

We will continue to strive to provide a more secure service. Thank you for your continued support!