AI Notice
What AI features do, their safety limits, and how AI-related data is handled.
1. You are interacting with AI
Responses in Ask AI and AI recommendation features may be generated or analyzed automatically by artificial-intelligence models. The interface identifies these features as AI so that users do not mistake them for a human conversation.
2. Accuracy limitations
AI can invent food or ingredient facts, misunderstand conditions, or provide inaccurate cooking or nutrition information. Verify important information using reliable sources such as product labels, manufacturers, public authorities, or qualified professionals.
3. Allergies and health
Ingredient exclusions and diet conditions tune recommendations; they do not diagnose or manage allergies or disease. If you have a serious allergy or health restriction, do not rely on an AI result alone to determine safety. Avoid entering unnecessary sensitive medical information.
4. Data processing
AI inputs and relevant conversation context may be sent to the OpenAI API to generate responses. Signed-in conversations are associated with the owning account so other signed-in users cannot continue them, and AI conversations associated with the account are deleted during account deletion. Conversation state is retained for 180 days after last activity; raw messages and recommendation presentation snapshots are retained for 180 days from creation and then automatically deleted.
5. Automated decisions
The current AI features are designed for food discovery and recommendations, not automated decisions that produce legal or similarly significant effects about a person.
6. Feedback
Report dangerous or materially inaccurate AI output to [email protected].