AI Use & Training Policy

Last Updated: April 13, 2026

This Policy explains how OX-san uses AI and how user inputs and generated outputs are handled.

1. How AI Is Used

We may use AI for recipe generation, meal suggestions, ingredient-use ideas, substitute ingredient suggestions, text summarization, and related features.

2. Handling of User Inputs

Ingredients, pantry data, requests, allergy information, and other content submitted by users are processed as needed to provide AI features. Users should not submit confidential information, another person's personal data, urgent medical information, or other content that is not appropriate to submit.

3. Model Training Policy

We do not use user inputs or user-specific AI outputs to train or retrain our own or third-party general-purpose models without the user's explicit consent. If we ever change this policy, we will provide clear advance notice and obtain any consent or complete any procedures required.

4. Use for Quality Improvement

For response-quality review, incident response, abuse prevention, and feature improvement, we may use anonymized information, statistical information, error logs, and other non-identifiable data to the extent permitted by law.

5. Limitations of AI

AI may generate incorrect information, including errors relating to ingredients, quantities, temperatures, cooking times, storage conditions, allergens, and nutrition values. Users should not rely blindly on suggestions and should verify package labels, public guidance, expert advice, and actual ingredient conditions.

6. Final Decision-Making

The final decision on whether to follow any suggestion, and whether to cook, consume, store, or serve food, always remains with the user.

Related Policies