AI results are estimates. Always check a fruit yourself before you eat it.
This AI Data Processing policy explains how Fruitly’s AI works, how your images are processed, how we protect your privacy, and the limits of AI results. It is part of our Terms of Use and should be read together with our Data Usage, Third-Party Access and Consent Settings policies.
Effective date: October 9, 2026
Last updated: October 9, 2026
Fruitly uses automated algorithms to identify fruit varieties and estimate ripeness for informational purposes.
When you scan a fruit, the AI analyzes features such as shape, color and texture and suggests the most likely match.
We currently use Google’s Gemini API to perform this analysis. We are also developing our own AI model.
Results are generated automatically. No person reviews your scan before you see the result.
Analysis (required): images are encrypted in transit and processed to recognize objects, color and texture. Scanning does not work without this step.
Improvement (optional): if you allow it in Consent Settings, de-identified scan data helps train our models. You can turn this on or off at any time.
Images are deleted after processing, unless you save them to your history or allow them to be used for improvement. See our Data Usage policy for details.
No biometrics: we do not use facial recognition or collect biometric data. Our AI is designed to identify fruit, not people.
Avoid capturing people: please do not include people, faces or personal documents in your scans. If they appear in the frame, the image is analyzed only to identify fruit.
Security: data is handled by trusted processors that are contractually prohibited from using it for their own purposes. We use Google’s paid Gemini API service, under which Google does not use your images to improve its products.
Data minimization: only the data needed for each scan is sent for processing.
AI results are based on probability and are not guaranteed to be accurate.
Ripeness results are estimates and may not reflect a fruit’s actual condition.
The AI cannot detect internal spoilage, contamination, pesticide residue, toxicity or allergens.
Results may be wrong for fruits that look alike, rare varieties, damaged fruit, or photos taken in poor lighting.
Related information, such as nutrition and storage tips, may be general, incomplete or incorrect.
You are solely responsible for checking the safety of any fruit before buying, eating or serving it.
Never eat a wild, foraged or unfamiliar fruit based on the App’s results.
Do not use the App for medical, allergy or dietary decisions.
You rely on AI results entirely at your own risk. Our Terms of Use, including the Disclaimer of Warranties and Limitation of Liability sections, apply.
If a result looks wrong, report it using the reporting tool in the App or by emailing fruitlyvibin@gmail.com. Your reports help us improve accuracy.
We may change the AI models we use, including when we move to our own model. We will update this page when we do. If a change significantly affects how your data is processed, we will notify you in the App or on this page and update the “Last updated” date.
Fruitly
[Business Address]
Email: fruitlyvibin@gmail.com