A public record
for the AI era.
Digital Fraud Observatory is a public record of how deception appears online: the claims, patterns and evidence that help people recognize fraud before it reaches them.
One project. Many sources.
Initiated by desenmascara.me, this project welcomes work from security vendors, researchers, regulators, educators and individuals. A source citation is not a partnership or endorsement. The collection contains 28 documented examples, including public incidents and clearly labeled fictional reconstructions. It does not imply an established contributor network.
What a case means
Each case identifies its source, dates, evidence type and limitations. A provider’s risk assessment stays attributed to that provider. This site does not run live website scans, give a fresh safety verdict or certify a business.
A screenshot can show a claim. It cannot, on its own, prove stolen funds, a coordinated campaign or AI generation. We keep those distinctions visible.
How review works
- Check the public report and the origin of its evidence.
- Separate visible facts, source assessments and editorial explanations.
- Check for duplicate cases, private information and unsupported accusations.
- Credit the source and contributor, then publish the approved case.
These are editorial checks, not a promise of independent incident investigation. Reports from all vendors follow the same process. The project does not publish an accusation simply because it was submitted.
Corrections belong in the record.
Use “Improve this case” to supply a correction and supporting source. When a correction changes the meaning of a case, its evidence, status and checked date should be updated together. Withdrawn cases should explain the withdrawal instead of quietly losing their history.
Built to be reused.
The original site code is MIT licensed. Original editorial explanations are available under CC BY 4.0. Third-party screenshots, report text, names and logos keep their existing rights. Download a case or the project files to inspect the structure.
Download the standalone project · Download case metadata · ASCII logo
A few practical questions
Can my company contribute its own research?
Yes. Link to a public report and explain the lesson. Your company is credited as the source. Disclose your connection and include evidence rather than promotional copy.
Does an AI-generated site mean it is a scam?
No. AI can be used legitimately. We document deceptive behavior and state AI involvement only when the evidence supports it.
Do I need to write code or use GitHub?
No. The contribution page prepares an email draft from a report link and a short explanation. GitHub issue templates are also included in the project files.