Start with the question the photo can answer

An image check can look for signs of generation or manipulation in the submitted file. It cannot establish the name of the person pictured, who controls the account or whether their story is true. A real photograph can be copied into a misleading profile; an AI-related signal can also be mistaken.

Keep those questions separate while you investigate. A detector result is one piece of information about the image, not a verdict about the person.

Keep a usable copy and check its context

Use the clearest version you can legitimately obtain. A small screenshot of a screenshot gives the analysis less useful information. Avoid adding filters or repeatedly saving the image before checking it.

A separate Google Lens search may show matching or similar public images. Compare the pages themselves, not just thumbnails. Finding no match does not establish that the photo is original.

  • Keep the profile link and the date you saw it for your own notes.
  • Compare details across the profile and conversation, without uploading private chats or identity documents to a detector.
  • Treat pressure to send money as a separate warning that a photo score cannot resolve.

What the DeepfakePolicy report adds

A standard photo check costs 1 credit. Read the findings, analysis limits and suggested next steps alongside the score. If the result is uncertain, preserve that uncertainty rather than forcing it into a yes-or-no answer.

From the completed report, you can add public-source research and context checks for 3 more credits. This extends the same report; it does not repeat the standard check. It may supply useful public references, but it is not a face-identification service or a guarantee of finding the original photograph.

Choose a next step you can verify independently

The FTC advises against sending money to a romantic contact you have not met in person and recommends checking profile images. If a finding worries you, pause and discuss it with someone you trust. Avoid publicly accusing someone on the basis of a detector score.

Download the report in your selected site language if you want to keep the findings. A saved report should help you ask better questions; it does not replace independent confirmation.

FAQ

Frequently asked questions

Can a photo detector tell me who the person is?

No. DeepfakePolicy checks image signals and optional public-source context. It does not identify the person in a photograph.

Does a low AI score make a dating profile safe?

No. A genuine photo can be reused by someone else. Verify the profile and the person's claims separately.

Automated results require source and context review.

Continue with independent verification.

Check the profile photo
Sources

Primary reading

We use original standards, regulators, public institutions and research papers wherever possible. Sources were last checked on 11 September 2026.