Use a newsroom workflow to verify the source, file history and AI signals in photos and videos before publication. Use AI Fake Scan to inspect the file, then corroborate who captured it, where and when.
Assign a reviewer and write down the exact claim the media is supposed to support. A copy can be genuine footage of an entirely different event.
Preserve the original and the post
Save the post URL, caption, account and time observed. Request the original file directly from the person who captured it and store the unedited copy under your newsroom’s access rules.
Trace the source and check the scene
Reverse-search the photo or several distinctive video frames. Compare signs, landmarks and weather with independent reporting, maps or other footage. Contact the source and record any gaps in the chain of custody.
Inspect what the file can tell you
Use the C2PA checker for a supported image’s signed edit history, or the watermark checker for supported providers. In a video report, read the frame and audio coverage separately. Passive visual findings remain Inconclusive in Basic Scan.
Record your evidence and decision
Log the source, claimed place and date, the corroborating links, what the scan checked and the open questions. Confirm permission to use the media and have an editor decide whether to publish, hold or escalate.
Check what the signer declared, which asset it applies to and what the checker validated. A valid signature supports the file’s declared history. Confirm the events in the scene independently.
Find the full recording, transcript or an independent recording. Basic Scan reports available audio-track information. Voice-clone classification requires a separate check.
Check your newsroom’s handling rules before uploading sensitive material. AI Fake Scan retains uploaded originals privately until you delete them through the scan page.