News Photo Verification for Journalists
Check whether a press photograph was AI-generated or altered before it runs. Free to try, and the file is deleted after analysis.
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
How every scan runs
Read the methodProvenance
C2PA content credentials and EXIF metadata, checked first
Watermark & Pixels
OpenAI and Google AI watermarks read, then a detection model scores the image
Evidence Chain
Every finding itemised and graded, not reduced to a yes or no
- Files deleted after analysis
- No account required
AI-Generated Disinformation in the News Cycle
FauxLens is built for journalists, photo editors and fact-checkers working to deadline.
Every result shows its reasoning, so a desk can see why a photograph was flagged rather than take a score on trust.
The volume of AI-generated imagery entering news pipelines has grown exponentially since 2024. Midjourney, Flux, and DALL-E 3 can produce photorealistic photographs of events that never occurred (floods, protests, political figures in compromising situations) in under 30 seconds. These images spread through social media before newsrooms can respond. Outlets have published AI-generated photographs that were later retracted.
The incidents that defined this problem share a single common failure: all were authenticated by eye rather than by forensic analysis. Human visual inspection is an unreliable test, and it does not get better as AI generators improve, because human judgment cannot measure the invisible mathematical properties that separate synthetic from authentic.
The most harmful incidents were not the obvious ones. They were the photorealistic fakes that passed editorial review: an AI-generated image of a burning neighborhood circulated as genuine wildfire documentation. A fabricated image of a political figure in a compromising situation spread to millions of viewers before correction. In each case, experienced visual journalists were relying on the eye alone, with no provenance check or trained detection model behind them. FauxLens adds that second opinion as support for human review, not a replacement for it.
How Forensic Photo Verification Works for Editors
FauxLens checks every uploaded image in stages, returning not a black-box verdict but a written evidence chain that an editor can evaluate and record.
Read moreShow less
Content credentials come first. FauxLens reads C2PA content credentials and EXIF/XMP metadata. A credential or generator signature from an AI tool ends the scan, and a genuine camera record can clear a photo. Missing metadata is never treated as evidence, because most social platforms strip it on upload.
Invisible watermarks come next. OpenAI and Google embed a hidden watermark (SynthID) in the images their AI tools create, and FauxLens reads it. When one is found, the report names the vendor family, not the specific model or prompt. Images from tools that do not watermark move on to the next stage.
A specialist third-party detection model then scores the pixels themselves, whether the image came from Midjourney, Flux, or any other generator. It judges whether the image is real, partly changed by AI, or generated by AI. The report turns that into one of four verdicts (AI-generated, likely AI-generated, likely authentic, or no AI detected), with an evidence line when only part of the image was changed.
When that model cannot return a result, a multimodal vision model reviews the image as a fallback, looking for visible signs of generation such as garbled text or objects that merge into each other.
Forensic examiners also use techniques such as Error Level Analysis (ELA), sensor noise (PRNU) matching, frequency analysis, and shadow geometry. FauxLens does not run those checks, so bring in a specialist when a story depends on them.
EXIF metadata is most useful on original files, so ask a source for the camera original whenever you can, since social platforms usually strip it.
The output gives editors a verdict, a confidence score (not a calibrated probability), and plain-language evidence lines, specific enough to inform an editorial decision and retain as documentation of your verification process.
Field Workflow Guide for Journalists
Building forensic verification into your daily workflow adds only a short step per image. Here is the step-by-step process.
Read moreShow less
When you receive an image from a source, wire service tip, or social media, upload it to FauxLens before any editorial use: before you start writing the caption, before you share it in Slack, and before you show it to an editor. The time a scan takes is small; the cost of publishing a fabricated image is not.
Review the verdict and the evidence lines behind it. The confidence score shows how strongly the evidence leans toward that verdict, but it is not a calibrated probability. AI-generated or likely AI-generated: treat as unverified and escalate to your photo desk. Low confidence: seek additional sourcing and request the original file. Likely authentic: continue your normal sourcing workflow. Do not publish without a chain of custody that traces the image to a credible source with a verifiable camera and location.
Cross-reference with reverse image search in parallel. Run a Google Images and TinEye search simultaneously with your FauxLens upload. Reverse search finds prior publication history - useful for detecting recycled images from unrelated events or stolen photos. It cannot detect brand-new AI-generated images that have never been posted before. For genuinely new content, provenance checks and a detection model are the tools that apply.
For wire services and photo agencies: request original raw files from your photo desk contacts when a FauxLens result raises flags. Original camera files keep the EXIF data and content credentials that social uploads strip, and they give a forensic examiner the sensor noise (PRNU) data needed for camera matching, a test FauxLens does not perform.
For newsrooms with high volume, FauxLens offers API integration that embeds directly into your CMS intake workflow. Your integration can route any image whose verdict is not likely authentic to human review before publication. Contact support@fauxlens.com for integration documentation.
Case Studies: AI-Generated Images That Fooled the Media
Several incidents from 2023 to 2025 define the stakes of photo verification failure. Each illustrates the visual tells and sourcing gaps that should have stopped the image.
Read moreShow less
The Pope Francis puffer jacket image (March 2023) was generated in Midjourney and showed the Pope wearing a large white puffer coat. It was shared widely on social media and reported as genuine by multiple outlets before fact-checkers traced it to Midjourney. Visible tells included a malformed hand around the coffee cup and eyeglasses that blended into the face. These are the kinds of rendering errors a detection model is built to catch.
The Pentagon explosion image (May 2023) showed a large explosion near the Pentagon building. It circulated on verified Twitter accounts and caused a brief dip in US stock markets before being debunked within minutes by geolocation analysis. Visually, the fence line blurred into the sidewalk and the building did not match the Pentagon's real facade, and no credible photographer ever came forward to claim the image.
The Trump arrest images (March 2023) circulated ahead of actual legal developments, generating significant media coverage of images that did not depict real events. The images showed the waxy skin texture, mangled hands, and blurred background faces typical of Midjourney output at the time.
The Maui wildfire images (2023) included AI-generated photographs of burning neighborhoods that were shared as documentation of the genuine disaster. The AI-generated versions showed the smooth, overly uniform fire and smoke rendering typical of diffusion model output, unlike authentic fire photography.
In each case, a provenance check and a detection model scan before publication would have given editors a reason to pause and verify.
Wire Services and Content Authenticity Tools
The major wire services have all begun implementing content authenticity verification as a structural part of their distribution pipelines, not as a reaction to specific incidents.
Read moreShow less
The Associated Press joined the Content Authenticity Initiative (CAI) and began implementing C2PA content credentials in its photo distribution workflow. Reuters launched a content authentication layer for its media library. Getty Images integrated C2PA verification into its contributor upload process.
C2PA, the Coalition for Content Provenance and Authenticity, is the technical standard underlying these initiatives. It embeds cryptographically signed metadata directly into image and video files at the point of creation. When a camera or software creates a file with C2PA credentials, the credentials record who created it, with what tool, and when. Because they are cryptographically signed, tampering with them is detectable. FauxLens reads C2PA credentials as the first step in its analysis. If a credential is present, it names the tool that made the file: an AI tool's credential ends the scan, and a camera record can clear the photo.
The first camera to ship with native C2PA signing was the Leica M11-P, released in late 2023. Several other camera manufacturers have announced C2PA support. Adobe Photoshop and Lightroom can add C2PA credentials to edited images, recording the editing history as part of the provenance chain.
The limitation of C2PA for journalism is that it covers content created by participating organizations with participating tools. Disinformation actors do not use C2PA-compliant tools. For content that C2PA does not cover, FauxLens checks for AI watermarks and has a specialist detection model score the pixels, which matters because unsigned images make up most visual disinformation.
FauxLens complements these institutional tools rather than replacing them. C2PA verification confirms authenticity of signed content. FauxLens detection helps screen synthetic content that was never signed at all.
Is that photo or video real?
Frequently asked questions
Photo forensics for journalists is the application of mathematical and computational analysis techniques to determine whether a photograph is authentic, captured by a real camera, or has been AI-generated, digitally manipulated, or composited. It goes beyond visual inspection to analyze pixel-level data including compression history, noise patterns, frequency domain signatures, and metadata.
Upload it to FauxLens. You will receive a verdict, a confidence score, and the evidence lines behind it. Also run a reverse image search in parallel to check prior publication history. If FauxLens returns anything other than a likely authentic verdict, treat the image as unverified and seek additional sourcing before publication.
Often. The detection model tells AI edits inside a real photo (such as generative fill or inpainting) apart from fully AI-generated images: an AI edit comes back as likely AI-generated, with an evidence line noting that only part of the image was altered. Edits made in Adobe tools usually carry a content credential that FauxLens reads directly. Traditional Photoshop retouching without AI is not flagged.
Yes, with limits. Social platforms usually strip EXIF metadata and content credentials, so on those copies the verdict rests on the watermark check and the detection model scoring the pixels. Accuracy can be lower on heavily compressed images, so request the original file when the stakes are high.
No, not on its own. A FauxLens result is supporting evidence for editorial decision-making and preliminary screening, not expert testimony or a certified forensic report. For legal proceedings requiring court-admissible expert testimony, the report serves as a starting point for engagement with a certified forensic examiner who can provide sworn testimony.
The Content Authenticity Initiative (CAI) is an industry coalition led by Adobe, along with the Associated Press, Reuters, BBC, and hundreds of other organizations, that develops and promotes the C2PA standard for embedding cryptographically signed provenance metadata in media files. When implemented, C2PA allows anyone to verify who created a piece of media, what tools were used, and whether it was modified - creating a tamper-evident provenance record from creation to publication.
The Associated Press, Reuters, and Getty Images have all implemented content authenticity programs. AP and Reuters use C2PA metadata verification in their distribution pipelines. Getty has integrated authenticity checks into its contributor upload process. These programs cover content from their own networks. For images sourced outside wire services (from social media, tips, and freelancers), independent checks such as reverse image search, contacting the source, and detection tools like FauxLens are the practical options.
If you find an AI-generated image being circulated as real news, report it to the platform where you found it using their content reporting tools. For images being used in news coverage, contact the outlet directly with your forensic analysis. Organizations such as First Draft, Full Fact, and Snopes accept tip submissions of suspected disinformation. Including the FauxLens verdict, confidence score, and evidence lines adds a documented detection result to your report.
Download the video using a Twitter/X video downloader tool and upload it to FauxLens. Note that Twitter/X re-encodes all uploaded video, which strips C2PA credentials. FauxLens samples frames from the video and checks them for signs of AI generation or editing, though accuracy on a re-encoded copy can be lower than on the original file.
As of 2026, most major platforms have policies requiring disclosure of AI-generated content in some contexts - particularly in political advertising. Enforcement remains inconsistent. Several platforms have implemented C2PA reading to display provenance labels on signed content. For unsigned AI-generated content, which represents the majority of visual disinformation, platform enforcement relies on user reports, automated classifiers, and third-party fact-checking partnerships rather than universal technical verification.
Keep exploring
Guides
- The Journalist's Toolkit: How to Verify Any Photo Is Real in 2026AI-generated images are flooding newsrooms. Here is the step-by-step verification workflow used by digital forensics professionals and investigative journalists to separate real from synthetic.
- The Science of Deception: How AI Detection WorksA detailed guide to the digital forensics of AI image detection, from Error Level Analysis (ELA) and JPEG compression artifacts to Photo Response Non-Uniformity (PRNU), and which checks a FauxLens scan actually runs.
- 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.
- Deepfakes in 2026: What They Are, Why They're Dangerous, and How to Spot ThemThe word 'deepfake' entered mainstream vocabulary. But most people still don't fully understand what it means or how dangerous it's become. This guide covers everything: the technology, the harms, and the detection.
More tools
Detect while browsing: try the Chrome extension
Right-click any image · No account required