How to spot a recycled video before you share it

2026-08-06 · wyper.io · misinformation, video, verification

Old footage posted as breaking news is the most common video hoax. Four checks that take under a minute, and what a fact-checker can and cannot tell you.

A video shows a flooded airport. The caption reads "SHOCKING WHAT'S HAPPENING IN DUBAI". The footage is real. The flooding happened. Everything in the clip is genuine, and the post is still misleading, because the event took place in April 2024.

This is the most common form of video misinformation, and it is the hardest to catch, because nothing in the frame is fake. The deception is in the framing.

Why recycled footage works so well

A doctored image can be caught by looking at the image. Recycled footage cannot, because there is nothing wrong with it. The claim attached to it is true in isolation: there was record rainfall, the airport was paralysed, people did die in the storms. Every individual statement checks out.

What does not check out is the implied "now".

Four checks, under a minute

1. Read the post date, not the videoEvery platform shows when a post was published. If the caption uses the present tense but describes a major event, ask why you did not hear about it yesterday.
2. Search the event, not the videoTake the most specific detail you can see, a place name, a building, a plane livery, and search it with the word that describes the event. Major floods, fires and crashes are covered within hours. If the coverage is two years old, so is your video.
3. Look for the original uploadReverse image search on a single frame often surfaces the earliest posting. The oldest copy is usually the real context.
4. Check what the account posts otherwiseAccounts that recycle footage rarely do it once. A timeline full of "BREAKING" with no follow-up is a pattern, not a coincidence.

What automated checking can and cannot do

An automated fact-check can compare two dates: when the event happened, and when the post was published. That comparison is mechanical and reliable, and it catches exactly this case.

What it cannot do is decide whether the poster meant to mislead. A journalist reposting archive footage with context and a rage-bait account are indistinguishable to a model. Intent is not a fact you can look up.

This is why a good check reports the time gap as its own finding, separate from the truth score, and leaves the judgment to you.

No. A truth score answers one question: is this claim supported by evidence. A post can carry a completely accurate claim and still mislead through framing, timing or omission. Read the score together with the context notes, not instead of them.

You usually cannot tell from the file. Metadata is stripped by most platforms on upload. What you can establish is when the event in the footage happened, by identifying something specific in the frame and searching for coverage of it.

Generally no, and it is often legitimate. Archive footage is used in reporting every day. What matters is whether the presentation makes a reasonable reader think they are seeing something current.

The short version

Real footage plus a false "now" is the cheapest hoax there is, and it works because our instinct is to check whether the video is fake. Check the date first. It is faster, and it catches more.