Old footage relabelled as new is the most common form of video misinformation. Learn how to verify the date and location of viral protest clips yourself.
This article was researched and drafted with AI, then checked and released by a person before it went live. The illustration is AI-generated. Every factual claim links to its source, so you can verify it yourself.
AI-generated illustration. A smartphone screen showing a paused video clip next to a calendar grid and a search bar.
To check if a protest video is old footage, you must locate its earliest documented publication online and compare it to the claimed event date. Reverse image searches of key frames can reveal if the clip appeared months or years before the current incident.
When a major political event or protest breaks out, social media feeds immediately flood with dramatic video footage. Many of these clips are highly compelling, showing massive crowds, intense confrontations, or shocking destruction. They also look entirely genuine, which makes them incredibly persuasive. The problem is that while the footage itself might be real, the caption attached to it is often a complete fabrication.
Recycling historical footage and presenting it as breaking news is currently the most prevalent form of video deception online. On September 27, the Times of India reported that Delhi Police had to issue warnings after a viral video claimed to show a massive, live protest against the Chief Election Commissioner at Jantar Mantar. The visuals were real, but they were taken from an entirely different, older event.
This tactic spans the globe. Just days earlier, Africa Check discovered that a viral video supposedly showing Kenyans burning the property of foreign migrants was actually old footage recorded in Uganda. The video was clear and unedited, but the context was entirely false. Similarly, Snopes investigated a viral September 2026 video claiming to show a recent interview about government benefits in Chicago, only to find the interview was actually recorded months earlier during the Democratic National Convention.
Because AI-generated media is dominating the headlines, as Al Jazeera noted in a recent report on election integrity, many users mistakenly assume that fake news always involves synthetic pixels. Relabelled video bypasses pixel detectors completely.
When someone takes a real video from 2021 and posts it today with a new caption, the file contains zero AI generation artifacts. The people are real, the background is real, and the camera movements are natural. If you run this file through a deepfake detector, the tool will correctly tell you that the video is authentic.
Furthermore, social media platforms heavily compress and process video files upon upload. As a recent industry guide on video editing explains, compression, stabilization, and automatic noise reduction can create odd visual anomalies that look suspicious but do not indicate deliberate human tampering. You cannot rely on visual artifacts alone to judge a video.
Instead of looking for fake pixels, you need to look for historical context. What carries the weight for verifying video is finding its earliest appearance online and measuring the gap between the claimed event date and the actual publication date. Catching relabelled footage requires a provenance check.
You do not need specialized software to perform a basic context check on a suspicious clip. The manual method relies on tools you already have in your browser.
First, pause the video on a frame that shows distinct details. Look for unique buildings, specific street signs, clear weather conditions, or prominent banners. Take a screenshot of this paused frame.
Next, upload that screenshot to a reverse image search engine like Google Lens or Yandex Images. These search engines will scan the internet for identical or highly similar images. If the footage went viral during an older protest, news organizations or fact-checkers likely published articles containing still frames from that exact clip.
If your search results bring up an article from three years ago featuring the same crowd and the same weather, you have successfully debunked the new claim. The footage cannot possibly depict today's events if it existed years in the past.
For a faster workflow, you can use our tools to investigate suspicious claims and media. The fact-checking features operate differently depending on whether you are verifying the text of a claim or the provenance of a video file.
If you want to evaluate the textual or spoken claims surrounding a video, you can use our standard text check. Fact-Check returns a truth score of 1-10 plus an evidence chain of real, linked sources. The standard check utilizes a Dual-AI cross-check powered by Gemini and Grok. Grok searches for its own information during this process, ensuring the two models do not share a single evidence set.
For users who need deeper research, our Full Spectrum check adds a Triple-AI reading. This involves three models evaluating one evidence set. The third voice does not perform a search. It receives exactly the sources the main model received and reads them independently. This means a disagreement is a different reading of the same material rather than a different search result. Our own published measurements show why this matters: in a test of 30 claims using frozen sources, Gemini and Grok agreed in 29 out of 29 cases. A third model of different origin disagreed in 3 of 26 cases, every time with a machine-verified quote.
When you need to analyze the video file itself, the AI Spotter tool provides a dedicated provenance report. This feature is available only in the web app. It is not a detector, and it never outputs a percentage of how artificially generated something looks. That percentage is the exact harm we exist to prevent. Instead, the AI Spotter checks for content credentials (C2PA) verified against 55 official trust anchors, file metadata, reverse image search results, earliest documented publication, and quote verification.
For video files, the AI Spotter takes still frames and searches each one separately. A hit on any frame counts as a hit on the entire video.
It is important to understand how social media affects content credentials. A content credential sits in its own separate box beside the picture data. Any repackaging of the file drops that box, even a lossless repackaging that leaves every pixel untouched. Because social platforms repackage files upon upload, a video taken off a feed will almost always return nothing on the cryptographic layer. The report states this clearly, as it is the normal case rather than a cause for suspicion. We always advise checking the original file instead, where the signature might still be intact.
When you want to evaluate the spoken claims within a clip, the fact-checking engine uses subtitles when they are reachable. If subtitles are unavailable, an AI model actually watches the video within a specific time window. The free tier covers up to 10 minutes on YouTube and 2 minutes on X. Pro users get 30 minutes on YouTube and 7 minutes on X. Longer videos are refused outright rather than checked partially.
Keep in mind that the web app cannot see your logged-in YouTube session, so subtitles are often unavailable there. The extension runs directly in your browser and can reach much more context. For videos on X, both tools take the same route.
| Feature | Manual checks | wyper web app | wyper extension |
|---|---|---|---|
| Cost | Free | Free tier for text, paid AI Spotter | Free tier available |
| Installation | None required | None required | Requires browser installation |
| Video watching | You watch and search | AI watches accessible URLs | AI watches in your active tab |
| Subtitle access | Manual extraction | Limited by YouTube session | Reads directly from the page |
| AI Spotter tool | Not applicable | Paid feature (4 credits/video) | Not available |
Who needs which: The manual route is best for users who want to learn the exact mechanics of reverse image searching and prefer not to use external tools. The free wyper web app is ideal for users on mobile devices or public computers who need to verify a specific claim quickly without installing software. The wyper extension is designed for heavy social media users who want to fact-check posts and videos directly inside their feeds without breaking their reading flow.
Social media platforms are increasingly experimenting with automated context labels. However, as research presented at the 2026 ACM Conference on human-AI complementarity highlighted, users do not just see an AI prediction in a vacuum. They see early judgments shaped by those indicators, which then influence how the information spreads.
Ultimately, no tool can replace human critical thinking. Our tools will never call a truth score proof of absolute trustworthiness, and they will never judge the intent behind a post. We provide the evidence chain and the timeline, but the final judgment always stays with the reader.
Sources are never filtered by origin in our evidence chains. There is no mainstream media bonus and no built-in preference for official sources. You receive the earliest documented publications and the verified quotes, allowing you to trace the claim back to its root.
By checking the provenance of a video before sharing it, you help stop the viral spread of recycled misinformation. The next time a dramatic protest video appears on your timeline, take a moment to look for its original source.
Relabelling old footage is the most common method of spreading false video claims online. By running a reverse image search on key frames, you can find the earliest documented publication and prove the video belongs to a past event. Always verify the timeline before you share breaking news clips.
The wyper Fact-Check extension runs these checks on the post itself: truth score, evidence chain, and the date gap that catches recycled footage.
The web app installs to your home screen in one tap. No store, no account.
Already using it? ★★★★★ Rate it on the Chrome Web Store. Reviews decide what others get shown first.
wyper