Where did it come from?
Provenance and metadata can link a file to its source. Missing provenance is a reason to look closer, not proof of a fake.
Content authenticity
Content authenticity for documents, images and video
Know whether a file is what it claims to be: where it came from, what changed, and whether AI made it.
The model
Three questions.
A clearer review.
What may have changed?
File forensics and generator traces show where a file may have been edited or made by AI.
Does the context support it?
A genuine file can still support a false claim. Compare it with the rest of the case.
From finding to next step
Make uncertainty useful.
A review should show what is known, what is still open and what to do next.
Illustrative review pattern · not an analysis result
| Observation | What it establishes | Next step |
|---|---|---|
| Original source is missing | The source is not yet established. | Request the original or a supported source record. |
| Two amounts disagree | The submitted values are inconsistent. | Check the itemization and supporting record. |
| A manipulation signal is raised | A supported check found a reason for review. | Inspect the finding and other available evidence. |
Our focus
Evidence your team
can work with.
Start with one recurring review, such as bank statements in a loan file or photos in a claim. We agree on formats, data handling and success measures before you share files.
How we think about evaluationFAQ
Content authenticity, answered.
What is content authenticity?
Whether a document, image or video is what it claims to be: that it came from the stated source, that it has not been edited since, and that it was captured rather than generated by AI. Trustable checks the file itself for edits, generator traces and metadata gaps, and shows each finding where it sits so a reviewer can decide what the file supports.
What are Content Credentials (C2PA)?
An open standard for attaching a signed record to a file that says which device or app made it and what was done to it since. Some cameras and editing apps attach them today; most files still arrive without one. Reading Content Credentials in Trustable is in development; until then, and for every file without a credential, the checks rest on the file itself.
Is AI-generated content always a problem?
No. An AI-made image isn’t fraud by itself; a product render, an illustration or a retouched portrait can be perfectly legitimate. What matters is whether the file supports the decision it is being used for. Trustable reports that a file shows generation or editing traces and where; your team decides whether that matters in context.
Does a detection score prove that a file is fake?
No. No AI is 100% accurate, and a strong signal is still a signal, not a verdict. Every finding comes with its location on the file, how the file appears to have been changed and how strong the signal is, so a person can weigh it against the rest of the case before acting.
Can I upload sensitive files through this website?
No. This website accepts no uploads. Book a demo and describe the workflow; we agree on data handling in writing before any file is shared, and the demo itself runs on a sample file. Files are then processed in the Trustable cloud, your VPC or on-prem; see security.
Content authenticity
See it on a sample file.
Book a 20-minute demo. We agree on data handling before you share anything.
Book a demo