Home / Topics / Digital & AI Readiness
Digital & AI Readiness · How-to
How to check if a video, photo, or voice is fake
Verify the person and the source. Looking harder at the picture is the least reliable check.
The honest answer: you often can’t tell by looking. The FBI says it can be difficult to identify when content is AI-generated.[1] So the reliable checks are about the source and the person, not the pixels.
Five checks that work
- Slow down when it’s urgent and emotional. Ready.gov warns about messages that ask you to act right away or sound too good to be true: think before you click.[2]
- Verify the person on a second channel. Hang up and call the number you already have, or look up the company’s real number yourself.[1]
- Find the original source. Search for the headline or image elsewhere. A reverse image search can show where a picture first appeared. Real news is usually reported by more than one outlet. (This tip is ours, not an agency’s.)
- Check Content Credentials when they exist. The free Content Credentials Verify tool shows provenance data some cameras and apps attach. Many files have none yet: credentials are still rolling out.[3]
- Ask something only the real person knows, or use your family code word.[4]
Visual hints (hints only)
The FBI lists things that can give away fakes: distorted hands or feet, unrealistic teeth or eyes, irregular faces, odd glasses or jewelry, wrong shadows, watermarks, lag, voice that doesn’t match, and unrealistic movement.[1] Treat these as reasons for suspicion, never as proof something is real. Better tools keep removing them.
What about AI detector apps?
Don’t treat a detector’s verdict as proof either way. The FTC has taken action against a company over claims about the accuracy of its AI content detection product.[5] Use the five checks above instead.
For organizations
NSA, the FBI, and CISA published an information sheet on deepfake threats to organizations, with steps to prepare for, spot, and respond to them.[6]
Sources
- FBI Internet Crime Complaint Center (IC3): PSA I-120324-PSA: Criminals Use Generative Artificial Intelligence to Facilitate Financial Fraud (3 Dec 2024)
- FEMA / Ready.gov: Cybersecurity
- Content Credentials (industry tool, not a government source): Content Credentials Verify
- Electronic Frontier Foundation (nonprofit): Worried About AI Voice Clone Scams? Create a Family Password (Jan 2024)
- Federal Trade Commission: Prepared statement to the Joint Economic Committee on the rising scam economy (25 Mar 2026, PDF)
- CISA: NSA, FBI, and CISA Release Cybersecurity Information Sheet on Deepfake Threats (12 Sep 2023, archived)
Official pages checked 16 November 2026. Guidance changes; follow your local officials first. How we research.
Page checked on · Change log · Report an error · Calm by design