Chad PSL AI18+
Privacy and honesty

How we’re different

What we measure, what the AI is and isn’t allowed to do, what we never keep, and the rules every screen follows. Only things that are true of the app today.
Updated
  • Front and side photo
  • 478-point face mesh
  • Photos not kept on our servers
  • Writing found in your photos is blurred before the AI sees them
  • Adults 18+ only, with an automated age check
  • Sources shown on every plan step
  • No credit used on a rejected photo

Two photos, measured

A front photo measured on a 478-point face mesh, and a side photo whose outline we trace, so the profile is measured rather than guessed. A measurement a photo can’t show fairly (a turned head, a beard over the jaw, thick frames) is left out and said so.

The AI rates a rubric, never your face

The AI model reads your photos against fixed, written rubric items (skin clarity, how well-kept your hair is…), twice, and the two reads are merged. It never gives your face a score and never writes your tier. A deterministic engine (version 7.8.2) turns the measurements and the rubric into the seven categories and the overall: the same inputs give the same numbers.

Photos never stored on our servers

Your face geometry (about 478 points) is computed in server memory for one request and is never written to our disk, database, logs or backups. Uploaded photos are deleted from our systems as soon as your analysis finishes, fails or is rejected. We keep scores and rounded measurements, never the photo or the mesh.

Writing in photos is blurred first

Before the AI sees a photo, our vision service finds writing anywhere in it (a sign, a caption, a sticker, a card) and blurs it, so text in a photo can’t steer the read. Writing over the face itself means a retake.

A real 18+ gate

You confirm you are 18 or older before any scan, every photo pair goes through an automated age check, and a photo that reads as a child is refused and pauses scans on the account and device for 30 days. The app is set for adults only in both stores.

Honest by rule

No percentiles or "top X%" about people: the 1–8 scale is calibrated on a modelled set of faces. No before/after faces, no celebrity comparisons, no surgery or filler suggestions, no folklore like mewing except as "skip, and why". Potential is an estimate, never a promise.

Every plan step shows its evidence

Each step in your plan carries an evidence grade, what it costs, when you could tell, who should skip it and its sources. Style suggestions (haircuts, beards, frames) are labelled as such and never change a score.

What happens to a scan, step by step

  1. You take a front and a side photo in the app. It guides your distance and pose as you frame each shot.
  2. The photos go to our server, where they live in memory only. The vision service maps the mesh, traces the profile and blurs any writing it finds.
  3. Photo checks run first: one face, pose, light and writing over the face.
  4. The AI model reads the blurred photos against the written rubric, twice, and checks apparent age; the reads are merged, and a photo that reads as a child is refused. A rejected photo never uses a credit.
  5. The engine scores the seven categories, the overall, the potential and your tier, and builds the plan.
  6. We store the scores and rounded measurements. The photos and the mesh are gone from our server.

The policies, in full

The biometric data policy, the privacy policy and the terms are the binding versions; this page summarises them.

Questions people ask

Does my photo leave my phone?

Yes, to be analysed: it is sent to our server, held in memory for that one request, and deleted when the analysis finishes, fails or is rejected. OpenAI, which rates the photos for us under its no-training terms, may keep a copy for up to 30 days for abuse monitoring, then deletes it. Your photos and the front photo’s mesh also stay on your own device for up to 24 hours so the analysis can be finished, then they are deleted.

Does the AI decide my score?

No. The AI rates fixed rubric items (skin, hair, grooming, parts of the profile) and checks the photo; it never scores your face and never names your tier. The scoring engine combines those reads with the measurements, and the tier is read off its number.

Is my data sold or used for training?

We never sell, lease, trade or commercialise biometric identifiers. OpenAI rates the photos under its no-training terms.

Why is there no web version of the scan?

The app is mobile-only by design: the age check, the camera guidance and the on-device handling are built for the iPhone and Android apps. This website has guides, a sample report and tools that don’t score anything.

Can I delete everything?

Yes. Delete your account in the app (Account → Privacy and data → Delete account) or on the web through the account deletion page. It removes your profile, email, scan reports, consent records and credits. Age-check records, which hold no photo or score, are kept for the time the privacy policy states.