Is This Real? A Practical Way to Check Claims, Images, and Screenshots
About this check
“Wait… is this real?!” Start with the specific check
If you have just seen a claim, image, or screenshot, the useful question is not simply “Is this real?” It is: What exactly am I trying to verify, and what would count as an answer?
For a written claim, a finished check should tell you whether the claim is true, false, or inconclusive, along with the evidence used to reach that verdict. For an image, the answer should indicate whether it is likely authentic, likely AI-generated, or unverifiable. A screenshot should be checked for the text it contains as well as the context around it.
That gives you a practical starting point: copy the exact claim, preserve the image or screenshot, and look for sources that can support or contradict the specific thing being asserted. Do not begin by trying to prove a broad feeling such as “this looks suspicious.” Begin with one checkable statement.
The manual method
1. Separate the claim from the presentation
Write down what the post is actually saying. A dramatic caption, a familiar logo, or a screenshot of a conversation can make a message feel authoritative, but those details are not the claim itself.
Turn the post into a sentence you can check. Include the person, event, place, or date if those details are part of the assertion. If the image is the subject, note what you are trying to establish: whether it depicts a real event, whether it has been altered, or whether it may have been generated.
2. Preserve what you saw
Keep the original wording, image, or screenshot available while checking it. Text inside a screenshot may differ from the caption that accompanies it. A cropped image may also omit information needed to understand what it shows.
If you cannot identify the exact wording or the relevant visual detail, mark that uncertainty rather than filling in the gap yourself.
3. Search for the claim in precise language
Use the distinctive words from the claim and look for reporting, records, statements, or other source material that addresses that same assertion. Compare what each source actually says with what the post says. A source discussing a related event does not automatically establish the claim in front of you.
Pay attention to disagreement. If the available evidence supports the claim, record that. If it contradicts the claim, record that instead. If the evidence does not settle the question, the honest result is uncertainty—not a forced yes or no.
4. Check the image separately
An image can circulate with an incorrect caption, and a screenshot can be presented without the conversation or page it came from. Check the visible text, the stated context, and the connection between the image and the claim.
Manual checking can stop when the original context is missing, the available material does not establish when or where the image came from, or the evidence remains conflicting. At that point, “unverifiable” or “inconclusive” is a more accurate answer than confidence built from appearance alone.
Where manual checking stops working
Manual verification depends on having a precise claim and enough accessible evidence to compare. It becomes difficult when a post is vague, an image is detached from its original context, a screenshot contains the key information, or the available sources do not resolve the question. It also takes judgment to keep the conclusion proportional to the evidence: a result can remain unsettled even after you have found several related pages.
That contrast matters. Checking a claim is not the same as finding something that sounds similar. A finished answer needs a clear verdict, the confidence attached to it, and sources that let you inspect the basis for the conclusion.
A direct way to check it
VOM verifies claims using real-time web search and cites the sources behind each verdict. For a checked claim, it returns a verdict, a confidence score, and the sources it used. Its verdict can be true, false, or inconclusive, so evidence that cannot settle a claim is reported as inconclusive rather than forced into yes or no.
For images, VOM assesses whether an upload is likely authentic, likely AI-generated, or unverifiable. It also reads the text inside a screenshot, so you can check a forwarded image without retyping its contents.
The useful question remains the same: what, exactly, are you asking to verify? State that clearly, inspect the result and its sources, and treat an unresolved answer as unresolved.