How it works

Three layers of detection. One clear answer.

FakeFind Pro combines AI-generation detectors, metadata analysis, and scam-pattern recognition to give you a plain-English verdict in under a minute.

Try it now
Uploaded content flowing into a scanning engine for analysis

From upload to answer

1
Upload or paste

Drop a file, paste an email, or share a link. No account required.

2
We analyse

Your content runs through three detection layers in parallel on our GPU-backed servers.

3
Get a clear answer

A 0–100 risk score, plain-English explanation, and a recommendation you can act on.

The three detection layers

Every submission is checked against multiple signals. We combine them into a single confidence score and show you which signals fired.

AI-generation signals

Specialised models look for the fingerprints of diffusion image generators, voice-cloning tools, and large language models in your content.

Metadata & provenance

We inspect EXIF data, codec patterns, C2PA content credentials, and known reverse-image matches to verify where the file actually came from.

Scam-pattern recognition

Text models check for urgency cues, lookalike domains, invoice-redirect patterns, and other social-engineering hallmarks.

Reading your result

Every check returns a single 0–100 score plus a risk band. We always show our reasoning — a score without explanation isn't useful.

Low risk · 0–39

No strong signs of AI generation or scam patterns. Still verify high-stakes decisions through a trusted channel.

Uncertain · 40–69

Some suspicious signals. Treat with caution — confirm sender identity or content source before acting.

High risk · 70–100

Strong signs of AI generation or known scam patterns. Don't send money, click links, or share data.

AI likelihood & scam likelihood

Alongside the Trust Score, we show two simple bands so you can tell what kind of risk you're looking at.

AI likelihood

How likely the content was made or changed by AI tools.

  • Low — looks consistent with a real photo, voice, or human-written message.
  • Medium — some signals look synthetic. Worth a closer look.
  • High — multiple strong signs of AI generation or voice cloning.
Scam likelihood

How closely the content matches known scam patterns.

  • Low — no clear match to known scam scripts or fraud patterns.
  • Medium — some scam-like language or sender cues. Slow down.
  • High — closely matches well-known scam scripts. Don't pay or click.

What we check, by content type

High level, in plain English — no model names, no jargon.

Voices

We check voice notes and calls for signs of AI voice cloning — unnatural breathing, missing micro-pauses, and synthesis artefacts. We also compare the script to known 'family emergency' and impersonation patterns.

Messages & emails

We read message and email text for scam scripts (romance, invoice, CEO fraud, phishing), urgency cues, lookalike sender domains, and reply-to mismatches.

Images

We look for fingerprints of AI image generators, missing or odd metadata, repeated patterns, and manipulation signs like cropping and re-encoding.

Video

We sample frames and audio and check both — for deepfake artefacts, lip-sync mismatches, and inconsistencies between the visual and audio tracks.

Limits & responsible use

  • • No detector is perfect. A Trust Score is a second opinion, not a verdict.
  • • Heavy editing, screenshots, and re-encoding can hide signals — treat uncertain results as uncertain.
  • • For high-stakes decisions (money, identity, safety), always confirm through a trusted channel.
  • • Don't use FakeFind Pro to harass, dox, or 'expose' anyone. Use it to protect yourself and your family.

Privacy by default

  • • Uploads go to private storage. Access is granted only by signed URL.
  • • Originals are deleted on a regular schedule once analysis is complete.
  • • We never sell your data and we never share files with advertisers.
  • • Anonymous checks are accessible only via a private share link you control.
Read our privacy policy

What you can check

Frequently asked

How accurate is FakeFind Pro?

No detector is perfect. We combine multiple signals — AI-generation patterns, metadata, provenance markers, scam-text features — and return a confidence score so you can weigh it against context.

What content can I check?

Email and message text, images (JPG/PNG/WEBP), voice notes and audio (MP3/WAV/M4A/OGG), short videos, and links to social posts.

Do I need an account?

No. Anonymous checks work for everyone. Sign up to see your history and share private results.

What happens to my uploads?

Files are stored in private, access-controlled storage and used only to produce your result. Originals are deleted on a regular schedule.

Can I forward emails to be checked?

Yes — forward suspicious messages to check@fakefindpro.com and we'll reply with a summary and a link to the full result.

Is there an API?

We're working on it. Email hello@fakefindpro.com if you'd like early access for your team or product.

Try it on something suspicious.

Anonymous, free during beta, results in under a minute.

Check content now