Privacy Policy — FutureUs

Last updated: 8 October 2026

FutureUs ("the app") lets you upload a photo of two people and generates an image showing how that couple might look years later, or with a future child. It can also take a photo of your own child (alone or with a parent) and show how they might look as a teenager and as an adult. This policy explains exactly what happens to your data. It is written to match what the software actually does.

No account, no name, no email

FutureUs has no sign-up and no login. We never ask for your name, email address, phone number, or any other identifier that points to you personally.

When you first open the app it generates a random identifier (a UUID) that is stored on your device. It is not derived from your device, your phone number, or any advertising ID, and it cannot be linked back to you. On Android, if you uninstall the app, that identifier is lost and a fresh one is created on reinstall. On iOS, it is stored in the device Keychain, so it may persist across reinstalls.

Your photo

Your permission first. Before you choose your first photo, the app tells you where it will go and asks for your permission ("I agree" / "Cancel"). If you cancel, the photo picker does not open and nothing is sent. Once you agree, the upload screen shows "AI permission given · Withdraw permission"; withdrawing stops all future sending until you agree again. It does not delete anything already sent — to have that deleted, email us (see "Your choices").

When you tap the generate button, the photo you selected is sent over an encrypted connection to our server, which forwards it to our AI image provider, fal.ai. fal.ai runs Google's Gemini image model to produce the result.

What we keep:

At our provider. We instruct fal.ai not to store the request that contains your photo, and to delete the files it generates after 24 hours. Google processes the photo under its paid API terms: it does not use it to train its models, and may keep it for a limited period for abuse monitoring under its own terms. For "Who do they look like?", OpenRouter only relays the request from fal.ai to Google; its handling is governed by its own terms.

Face detection runs on your device before anything is uploaded. Each frame expects a fixed number of faces (two for a couple; one, two or three for the "our child" frames, depending on who you say is in the photo). If the photo does not contain that many faces, no upload and no generation happens at all.

Photos of your child. The "our child, grown up" frames take a photo of a child. Everything above applies to that photo exactly as it does to a photo of a couple: it is not stored, only its one-way fingerprint is kept, and the result link is deleted after 24 hours. Upload only a photo of your own child. The app is used by the parent; it is not directed to children (see below).

HD upgrade. If you upgrade a frame to HD, the generated frame is sent from your device to our server and on to the same image-generation provider to be enlarged. It is handled exactly like an uploaded photo: not stored, only fingerprinted, with the result link deleted after 24 hours.

"Who do they look like?" You can ask which parent your child takes after. You provide either one photo showing mother, father and child together, or three separate photos (mother, father, child). Face detection runs on your device first (three faces, or one face per photo); if it does not match, nothing is uploaded. The photos are then sent from your device to our server and on to a vision-language model (Google's Gemini, reached through fal.ai and OpenRouter), which returns only numbers and labels — a percentage for each parent and a short list of features such as "eyes" or "chin" — and is instructed not to comment on skin tone. The photos are handled exactly like an uploaded photo: not stored, only fingerprinted. We keep the returned numbers and labels for 24 hours, like a result link, and delete them automatically. This is an AI reading for fun. It is not a prediction, not a genetic or biometric assessment, and must not be used as one. Upload only photos of your own family.

Video. Turning your frames into a short video happens entirely on your device. Nothing is uploaded for it.

Face data

This section describes everything FutureUs does with faces, in one place.

What face data the app uses. FutureUs does not create, collect or store face geometry, facial landmarks, "faceprints" or any other biometric identifier, and it does not recognise or identify anyone. The face-related data it handles is:

  1. On-device face detection. Before any upload, the app uses Apple's Vision framework (Google ML Kit on Android) on your device to count the faces in the photo you chose and to score how clearly each face was captured (a number from 0 to 1 for blur, size and angle). The face positions and scores are used only to decide whether the photo is suitable, and are discarded right after that check. Face positions are never uploaded.
  2. A technical summary of that check. To tune the check, the app sends one usage event with: whether the photo passed, the number of faces found, the lowest quality score, the threshold applied, and the photo's pixel count. It carries the anonymous identifier described above and no image or face position. It goes only to our own server — never to an analytics company or any other third party — and is used only to set the photo check.
  3. The photos you choose to upload, which contain faces. They are processed as described in "Your photo" above.

How it is used. Only to produce the result you asked for: the aged or "future child" image, the HD upgrade, or the "Who do they look like?" reading. Face data is not used for identification, authentication, tracking, advertising, profiling, or to train any AI model, and is never sold.

Who it is shared with. Uploaded photos are sent to fal.ai, our AI image provider, which uses Google's Gemini models to generate the result. For "Who do they look like?", fal.ai passes the photos through OpenRouter to Google's Gemini model, which returns only percentages and feature labels. These companies process the photos on our behalf to return the result; we do not share face data with anyone else.

Where it is stored and for how long.

Your permission. The app asks before the first photo is chosen, and you can withdraw at any time with "Withdraw permission" on the upload screen; after that nothing is sent until you agree again. Withdrawing stops future uploads; it does not delete data already sent. To delete that, email us.

Deleting it. Uninstalling the app removes the photos and results kept on your device. On iOS, the anonymous identifier may remain in the Keychain (see "No account" above). To delete the fingerprints and usage events linked to your device identifier, email us at the address below.

Usage data

To understand how the app is used and to improve it, we record events such as: app installed, session started, a result generated, a paywall shown, a purchase completed. Each event carries technical context: app version, platform, OS version, device model, language, store country, and the anonymous identifier described above.

We do not collect your location, contacts, calendar, photo library contents beyond the single photo you choose, or advertising identifiers.

Purchases

Subscriptions and frame packs are processed by the App Store (Apple) or Google Play, depending on where you installed the app. We never see or receive your payment card, billing address, or Apple ID / Google account details. We use Adapty to verify purchases and manage subscription state: Adapty receives the store's purchase receipt, the product you bought and the app's anonymous identifier, and tells us whether that identifier has an active subscription, or bought a frame pack, and how many frames it is entitled to — nothing more. Frames from a pack are kept as a number against the anonymous identifier for 12 months from purchase. Advertising-identifier and IP-address collection is switched off in Adapty's SDK.

Where your data is processed

Our server and database run on Supabase infrastructure in the United States (us-west-2). Image generation is performed by fal.ai using Google's Gemini models; for "Who do they look like?", fal.ai reaches Gemini through OpenRouter. Subscription state is managed by Adapty. Each of these providers processes data on our behalf.

Children

FutureUs is not directed to children and we do not knowingly collect data from children. The app is for adults (18+). Some frames let a parent upload a photo of their own child; that photo is handled like any other — never stored, only fingerprinted — and no account or profile is created for the child. If you believe a child has used the app, or you want a fingerprint derived from a photo of your child removed, contact us at the address below.

Your choices

Changes

If we change what the app collects, we will update this page and the "last updated" date above.

Contact

E-mail: info@sentovex.app