Fake Screenshot Detector
Someone sent you a screenshot and you want to know if it is real. Upload it and FauxLens checks whether it was made or changed with AI. Free to try, no account, and the file is deleted after the check.
Upload the image or video. It is deleted after the check.Drop a photo or video here, or click to upload. It is deleted after the check.
JPG PNG HEIC · MP4
Detects content from
- OPENAI
- MIDJOURNEY
- STABILITY
- GEMINI
- FIREFLY
- FLUX.1
- CANVA
- IDEOGRAM
Key facts about this screenshot check
How it works- Checks the file’s content credentials and metadata, the invisible watermark OpenAI and Google put in images their AI tools make, and the pixels, with a detection model.
- Takes a photo or a video: JPG, PNG, WebP, HEIC, MP4 or MOV.
- Gives a verdict, a confidence score and the evidence behind it.
- Your file is deleted after the check.
- A page edited in the browser and then captured is a genuine screenshot, so no image check can catch it. Ask for the original conversation or a screen recording.
- Free credits while available. More come in one-time packs, no subscription.
What a result looks like
Every result comes with a confidence score and the evidence behind it. A low score means inconclusive, not clean.
AI-Generated or Likely AI-Generated
The checks found signs that AI made the image. If the evidence says only part of it was changed, the rest may be a real photo.
Do not act on it. Ask for the original, or check the claim another way.
No AI Detected or Likely Authentic
None of the checks found signs of AI generation in the image.
How the analysis works
A screenshot is not a photograph, and the forensics are different. FauxLens reads any AI watermark or content credential in the file and scores the pixels with a detection model for AI generation and AI editing, then returns a verdict, a confidence score and short evidence lines. It works best on the original file.
Bring photo-forensics habits to a screenshot and you will reach the wrong conclusion, because the two file types fail in almost opposite ways. Photos carry EXIF metadata (make, model, lens, GPS, shutter speed) that can be checked against known cameras. Screenshots carry none by design, and there is no camera sensor noise to fingerprint either, so missing metadata on a screenshot is expected, not a sign of tampering. The file is usually a PNG or a re-compressed JPEG.
AI generation is where FauxLens is strongest. A screenshot made by an image generator such as ChatGPT or Gemini can carry the invisible watermark OpenAI and Google embed, and the detection model scores the pixels for AI generation.
An edit made by hand in an image editor is different. When someone edits a JPEG screenshot and re-saves it, the edited region has been compressed twice while the surrounding interface has been compressed once. A forensic analyst can look for that trace, but the FauxLens scan does not flag edits made without AI, and it does not show where an edit is.
Screenshot forensics matters because screenshots are the main form of digital evidence in personal disputes, online commerce and social media conflicts.
How fake screenshots are made
There are four ways to fake a screenshot, and they are not equally detectable. Knowing which one you are probably looking at tells you how much weight a scan result can carry.
| How it was made | Can FauxLens catch it? |
|---|---|
| Edited in the browser (Inspect Element) | No |
| Edited in Photoshop or GIMP | Only edits made with AI |
| Made with a fake-screenshot generator app | Rarely |
| Generated with AI | Usually |
Read moreShow less
- Edited in the browser: the forger opens a live web page, such as a banking portal or a messaging app in the browser, right-clicks the text they want to change, picks "Inspect Element" and edits it in the page source. The browser renders the change instantly and they take a screenshot. It needs no image-editing skill, and the result is a genuine screenshot of a real browser showing fake data. A sharp eye may spot a font or layout that does not quite match the real platform.
- Edited in Photoshop or GIMP: a real screenshot is taken, then specific text or numbers are changed in an image editor and the file is re-saved. This is the method most likely to leave traces in the image itself, because the edited region is re-saved on top of the original.
- Made with a generator app: websites and apps render a synthetic interface from fields the user fills in (sender, message, timestamp, balance). The output is a clean render rather than an edited capture, though a side-by-side comparison with a real screenshot sometimes shows wrong spacing or font weights.
- Generated with AI: an image model produces a picture that looks like a screenshot but never was one. This is the case FauxLens handles best: the detection model scores the pixels for AI generation, and images made with OpenAI or Google tools can carry an invisible watermark the scan reads.
The four situations we keep seeing
Most fake screenshots fall into four groups: edited chats, fake payment or bank screens, inflated social media numbers, and invented app reviews or analytics. Each tends to be made a different way, so the method table above tells you how much a scan can catch in each case.
Read moreShow less
- Chat fabrication: the most common category. Conversations are changed to show words someone never said, usually by editing the page in the browser before taking the screenshot, which takes under 30 seconds and leaves a genuine capture of a fake page.
- Financial fabrication: fake bank statements, payment confirmations and trading screenshots used to deceive sellers, partners or investigators. They are made either by editing real screenshots in an image editor, which can leave traces around the altered numbers, or with fake receipt generator websites. See the fake payment screenshot detector for that case.
- Social media fabrication: follower counts, engagement numbers, DMs and profile statistics altered to look more credible, typically for influencer fraud or in online disputes. A careful edit of a few digits may pass unnoticed.
- App screenshot fraud: fake app store reviews, invented customer testimonials and made-up app analytics sent to clients or investors. Entirely AI-generated screenshots of apps that do not exist are increasingly common here.
Three checks analysts run by hand
Forensic specialists who examine screenshots by hand look at three signals that photographs rarely provide. FauxLens does not run these checks, but you can try the second one yourself.
Read moreShow less
- Compression entropy: text and interface elements compress very differently from photographic content. A chat background is large areas of uniform colour, while text is high-contrast detail. When a screenshot has been edited, the spread of compression across the image changes in ways that can reveal the tampered region.
- Font rendering: operating systems draw text with sub-pixel hinting (ClearType on Windows, Core Text on macOS) that produces predictable colour fringes at character edges. A name or number changed by hand often has fringes that do not match the text around it. Zoom in on it and compare.
- Interface geometry: chat bubbles, buttons, timestamps and status bars follow strict sizes set by the platform. Fabricated screenshots often break them slightly: a bubble two pixels too tall, a timestamp in a marginally wrong font. These are invisible to the eye but detectable with template matching.
When a faked screenshot cost real money
A faked screenshot costs money when someone acts on it before checking the source: a business wires funds, a seller releases goods, a court or an insurer weighs a chat that never happened. These are the documented patterns, from business email compromise to fake posts.
Read moreShow less
- Business email compromise: in 2023 the FBI's Internet Crime Complaint Center received 21,489 business email compromise complaints, with adjusted losses over $2.9 billion. A common variant sends fake wire transfer confirmations to accounts payable teams to "confirm" that money has moved, which delays discovery of the fraud.
- Marketplace payment fraud: sellers on Facebook Marketplace, eBay and Craigslist are sent edited screenshots of Venmo, PayPal or Zelle transfers that never happened, and release the goods before checking their own account.
- Legal disputes: edited text conversations are submitted in civil and family cases to show agreements or admissions that never happened. A scan can help decide whether to challenge a screenshot and ask for the original device or a platform export.
- Insurance claims: claimants submit fabricated chats in which an insurer's agent appears to authorise coverage or promise a payout, to dispute a denied claim.
- Fake posts: screenshots of posts by politicians, celebrities or executives are edited to show statements never made, then shared as "evidence". They do the most damage because the harm lands before anyone checks.
Is that photo or video real?
Related checks
Frequently asked questions
Upload the screenshot above and FauxLens returns a verdict in seconds, with no account. Fake screenshots are made in three main ways, and they are not equally detectable: generating the whole thing with AI, or changing it with an AI tool, leaves traces the scan can find; editing the captured image by hand is not flagged; and editing a live web page in the browser and then capturing it produces a genuine, unedited screenshot of a fake page, so there is nothing in the pixels to catch. For that last case, treat a clean result as weak evidence and ask for the original source instead.
Not reliably. FauxLens checks whether a screenshot was made or changed with AI. An edit made by hand, such as a few changed characters, is not something it flags, and it does not show where an edit is. Upload the original file rather than a forwarded copy for the best result.
Chat screenshot detection is one of the most common use cases. FauxLens can flag chat screenshots that were generated or changed with AI. Text edited by hand in an image editor is not flagged. It cannot catch a chat edited in the browser with developer tools and then captured, because that is a genuine capture of a fake page.
FauxLens can flag financial screenshots that were made or changed with AI. The detection model can flag a statement that was made or edited with AI, but it does not point to the changed figures. A statement from a fake generator site, or one edited in the browser, looks like a genuine capture. For legal or compliance use cases, combine FauxLens analysis with direct account verification.
PNG, JPEG, and WebP are all supported. PNG is the native format for most screenshots and keeps the most detail, so upload the original PNG when you have it. In any format, heavily compressed, forwarded or very small images give weaker results.
It is free to try, with no account and no card. More scans come in one-time packs with no subscription.
Yes, and they frequently are. iMessage screenshots can be fabricated by editing the DOM of the iMessage web export, using Photoshop on a real screenshot, or using dedicated iMessage fake generator tools. They are not equally detectable. A Photoshop edit made without AI is not flagged; a DOM edit produces a genuine capture with nothing in the pixels to find; and generator fakes are clean renders, though their bubble sizes and padding sometimes differ from real Apple bubbles in ways you can compare by eye.
Courts increasingly require corroborating evidence alongside screenshot submissions. The most reliable verification methods are: requesting the original device and extracting the conversation through a forensic tool (Cellebrite, Oxygen Forensics) to compare with the screenshot; obtaining platform records directly from the service provider via legal process (subpoena to WhatsApp, Apple, or the relevant platform); and forensic image analysis to detect editing artifacts. A FauxLens result can help you decide whether to challenge a screenshot and request a device examination, but it is a screening tool, not expert testimony.
WhatsApp screenshots can be checked by FauxLens for AI generation and AI edits in the same way as any other screenshot. For definitive verification in legal contexts, the most reliable method is requesting the original WhatsApp export directly from the device using a certified forensic tool, which produces a structured data export that can be compared against the screenshot. Alternatively, WhatsApp messages can be subpoenaed directly from Meta in some jurisdictions.
Several categories of tools create fake screenshots: browser-based fake receipt and fake chat generators (readily found by searching for specific platform names plus "generator"), Photoshop and GIMP for manual editing, and increasingly, AI image generation tools prompted to produce interface screenshots. FauxLens is strongest on AI-generated screenshots, does not flag image-editor changes made without AI, and cannot reliably catch generator-site fakes, which are clean renders that look like genuine captures.
Sources
Updated
Keep exploring
Guides
- 5 Visual Tells: How to Spot an AI Image Without ToolsBefore you run a scan, check these 5 common mistakes AI models make. From topology errors in hands to subsurface scattering fails in skin.
- EXIF Metadata Explained: The Hidden Data Inside Every PhotoEvery photo your camera takes contains a hidden layer of data: camera model, GPS coordinates, time of capture, and more. AI-generated images usually lack this layer. Here's how to read it.
- Romance Scam Warning Signs: A Visual Checklist for 2026AI has made romance scammers more convincing than ever. FBI data shows over $1.3 billion lost annually. Here is the visual checklist every person on a dating platform needs before trusting a profile.
More tools
Detect while browsing: try the Chrome extension
Right-click any image · No account required