Check a photo before you trust it
Some AI-generated or edited images contain Content Credentials or other provenance information. When available, we check and report that information, along with the file’s metadata and classifier signals.
What gets checked
Some AI-generated or edited images contain Content Credentials or other provenance information. When available, we check and report that information.
- Provenance. Content Credentials or similar information, when the file carries any.
- Metadata. The camera make and model, exposure settings, date and any editing software the file records.
- Classifier signals. A likelihood, from a model trained on AI-generated images.
- Visual indicators. Not assessed by this tool. The result lists what to look at yourself.
The result appears here: the evidence found, how strong it is, and what these checks cannot tell you.
The results you can get
- Labelled as made by AI
- Probably made by AI
- Something has been edited
- Nothing obvious found
- Can’t tell
- Camera record checks out
Missing evidence is not evidence. WhatsApp and most social apps strip metadata from everything they pass on, so a forwarded photo almost never carries a camera record. That is why “no camera record” is reported as inconclusive and never as suspicion.
We never certify anything as genuine. The strongest positive here is “nothing obvious found”, and it is deliberately not reassuring. There is no green tick on this page and there never will be.
AI detection is probabilistic. A result cannot by itself prove that an image is real or AI-generated, and it cannot prove authenticity on its own. How we check