BeFoxPrivacy Policy
This policy describes how Agentic AI LLC ("we") handles information in the iPhone app "BeFox" (the "App").
1. Our approach
The App can be used without an account, and we never ask for your name, phone number, email address or postal address. Text you send for a check is not stored on our servers unless you agree to share it as a case. Text is sent to an external AI service only after the App has shown you what will be sent and to whom, and you have agreed.
2. Information that stays on your iPhone
- Screenshot images: text is read on your device. Images are never sent off your device.
- Conversation history: the exchange shown in the App is handled on your device.
3. Information we receive, and why
- Text for a check: text you type or paste, and text read from a screenshot (also sent when the confirmation screen is shown, so that links can be checked). Used only to assess scam risk and show you the result; not stored after processing. Before any of it is sent to the AI service, the App asks for your consent (see 5).
- Anonymous identifier: a random ID generated by the App and kept in your device's Keychain. Used to apply the daily free limit and to manage your subscription status. It is not linked to your name or any other identifying information. The ID may remain in the Keychain even after the App is deleted.
- Result records: for each check, the stage, score, confidence and the kinds of scam detected — never the message text. Used to provide and improve the service.
- Usage counts: used to apply the daily limit.
- Feedback: if you choose to send it — whether the result helped, a reason, a short note, the score of that check, and who received the message (you, family, etc.). Email addresses, phone numbers and postal addresses in the note are removed automatically before it is saved. Used to improve quality.
- Subscription information: the transaction ID, product and expiry we receive from Apple. Payment is processed by Apple; we never receive your card details.
- Connection logs: IP addresses and access times are recorded by our servers and our content delivery network for security and troubleshooting.
4. Sharing a case (only with your consent)
After a high-risk result, the App may ask whether you want to share the case to help prevent fraud. Only if you agree, we store an excerpt of the message with personal details (phone numbers, addresses, names, account numbers) removed, together with the score and the patterns detected, and use it to improve detection and as reference cases. A shared case is stored without your identifier. If you decline, nothing is stored.
5. Service providers
- AI service provider (Google LLC, United States): to assess scam risk, and only with your consent, the text being checked (text you type or paste, text read from a screenshot and confirmed by you, and follow-up questions) is sent through our servers to Gemini, an AI service provided by Google LLC (paid Gemini API service). Images are never sent. The App asks for this consent on a dedicated screen before your first check; without it, nothing is sent to the AI service. Google states that for its paid Gemini API services it does not use prompts or responses to improve its products, and logs them for a limited period solely to detect and prevent prohibited use and for legally required disclosures. We use its service having confirmed that it provides protection equal to or greater than this policy. Processing may take place outside Japan, including in the United States. If we change providers, we will update this policy and ask for your consent again in the App.
- Apple: App distribution and subscription payments.
- Cloudflare: content delivery and protection.
6. Disclosure to third parties
We do not sell or provide your information to third parties, except where required by law. We do not show ads, and we do not track you across other apps or websites.
7. Retention
Result records, usage counts, feedback and subscription information are kept for as long as needed to provide and improve the service. Because we do not link information to your identity, we may be unable to identify records belonging to a particular user.
8. Security
We use encrypted connections (HTTPS), restricted access and other measures to protect information against leaks and loss.
9. Changes
We may update this policy as needed. Significant changes will be announced on this page or in the App. The Japanese version of this policy prevails in case of any discrepancy.
10. Contact
- Operator
- Agentic AI LLC (Japan)
- Contact
- [email protected]