How to Verify a Photo Online
Verification is a sequence, not a single test, and the order matters more than the tools. Done right, most images resolve in under two minutes — usually at step one, and usually not for the reason people expect.
Step 1: Reverse image search first, always
Start here, not with forensics. The most common form of visual misinformation is not a generated image or an edited one — it is a genuine photograph with a false caption, recycled from another place or another year.
No forensic tool can catch that, because nothing is wrong with the file. Reverse search catches it in seconds by finding earlier appearances.
Use more than one engine — Google, Yandex, TinEye and Bing index differently, and Yandex in particular is strong on faces and locations. Our guide covers the differences.
Step 2: Read the metadata
Free, instant, occasionally decisive. Look for a camera make and model, capture settings, and whether the capture date matches the claim.
Also look for generator signatures — some AI tools write themselves into the Software field — and for editor names. Our EXIF viewer needs no account.
Remember the asymmetry: finding metadata is informative, finding none is not. Every messenger and platform strips it.
Step 3: Forensic analysis
Now check what the eye cannot: whether the image was generated, and whether any region was manipulated. This is where the detector earns its place, running twelve independent methods and reporting each separately.
Read the breakdown rather than the headline number. Several methods agreeing is the signal; one loud method with eleven quiet ones is not.
Step 4: Look at the picture properly
Cheap and often decisive. Check shadow directions against each other. Check reflections against the scene. Check text on signs, plates and uniforms. Check ears, teeth and jewellery on any face.
For a claimed location, compare against satellite or street imagery — building shapes, signage and road markings are hard to fake consistently.
Step 5: Trace the source
Often the most decisive step and the most skipped. Who posted it first? Does that account have a history, or was it created last week? Does anyone credible corroborate the same event from a different angle?
A photo with no traceable origin, from an account with no history, should be treated as unverified regardless of what any tool says about the file.
What each step can and cannot tell you
Reverse search finds recycling, not manipulation. Metadata finds provenance, when it survived. Forensics finds generation and editing, not context. Visual inspection finds obvious errors, not subtle ones. Source tracing finds intent.
None of them alone is verification. Together, they usually are — and when they disagree, that disagreement is the finding.
Key points
- Reverse image search first: recycled real photos are the most common fake.
- Metadata is free and fast; its absence proves nothing.
- Forensic analysis covers what the eye cannot reach — read the breakdown, not the score.
- Compare shadows, reflections and text; check locations against satellite imagery.
- Source tracing is the most skipped and often the most decisive step.
- No single step is verification. Convergence is.
Frequently asked questions
Check an image yourself
Reading about it only goes so far. Our AI image detector runs twelve independent methods and shows you every score, not just a verdict.
More from the blog
Try it on your own image
Free account, 3 checks included, no card required.