Catfishing and Synthetic Media: What Evidence Can Show

Reviewed September 30, 2026.

Synthetic Media Changes The Questions

A suspicious profile may use a stolen photograph, edited media, a synthetic image, or a real person's image without permission. Keep the media-authenticity question separate from who operates the account and whether its conduct is fraudulent.

Avoid Visual Certainty

Odd lighting, distorted backgrounds, or inconsistent movement can justify closer review, but compression, camera conditions, and ordinary editing can also affect an image. Their presence is not proof of AI generation, and their absence is not proof of authenticity.

NIST's synthetic-content overview describes multiple approaches, including provenance and detection, with limitations. Record the detector, version, input quality, and stated uncertainty if a tool is used. Do not treat a detector score or a video call as conclusive identity verification.

Corroborate Beyond The Media

Preserve the source URL, available timestamps, and how the material was obtained. Review whether public names, profile history, and independently checkable claims are consistent. An empty image-search result only describes that search's coverage; it does not identify the origin of the image.

Writing style and posting times may have ordinary explanations, including translation, scheduling, or shared account management. Avoid attributing authorship or wrongdoing from those patterns alone.

Preserve Evidence And Limit Contact

Keep relevant messages already received and document requests for money or sensitive information. Use Verify Identity to discuss a permitted review if the evidence needs corroboration. Do not confront the account or send tracking links as a test. The useful output is a clear record of supported findings, contradictions, and unanswered questions.