VARNIUMAI risk defense / everyday bearings

DEEPFAKE / IMAGE CHECK

Is This Photo AI-Generated? A Check for Listings, Profiles, and News.

For a rental listing, a dating profile, or a news image, the question is not "is this AI" but "can I trust what this photo is being used to claim." The check reads purpose and provenance.

Start here

The free tool below runs entirely in this browser — no account, nothing stored or transmitted. Redact anything you would not want kept before you type it.

Open the companion tool →

Why pixels are a losing game

Generated images no longer reliably show the old tells, and real photos can be manipulated, cropped, or reused out of context. Pixel inspection gives false confidence in both directions.

The durable check is purpose and provenance: what claim is this photo being used to support, and is there an independent way to confirm that claim — a reverse search, a video call with the person, a visit to the listing?

The protocol

  1. Describe the photo and what it is being used to claim (a rental, a person's identity, a news event).
  2. Read the purpose reading: the claim, the source, and the risk if the claim is wrong.
  3. Run the independent confirmation: reverse-search the image, or request the live proof the claim would require.
  4. If the photo was used to take your money, contact your payment service and report at ReportFraud.ftc.gov.

What the tool does

Describe a photo and its claimed use and get a purpose-and-provenance reading with the basis shown. It runs in this browser; nothing is sent or stored.

Limit first: This check can raise your confidence, but a clean result doesn't prove a photo is real. It does not analyze the image — it structures what the photo is being used to claim.

Open the companion tool →

One human next step

Reverse-search the image and request live confirmation of the claim it supports; if money moved, report at ReportFraud.ftc.gov.

Research log and safety checks

AI-assist path: image-forensic models exist for authenticity signals, but are not wired into this static artifact; the checker structures purpose-and-provenance questions the reader can run now, and the gap is named. §6.10: not a frontier-pace topic.

§0.5 protection result: category is deepfake-media-verification; crisis-adjacent: no; fear/urgency/scarcity toolkit used: no. The protective function is free and the paid feature is convenience only.

Related dispatches

Pairs with the same protection bar: a free complete reading, an optional convenience unlock, and no fear-framing.