3 free checks to spot an AI-made image
No single check proves an image is real or fake, but three free ones give you real evidence. Google's About this image shows where else a picture appeared and when it was first seen. Gemini can look for Google's SynthID watermark. The Verify tool reads Content Credentials, if the file carries them. Use all three together.
What you need
| Apps | The Google app, Gemini and a web browser |
|---|---|
| Cost | Free |
| Time | About 5 minutes |
| You need | The image file, or a screenshot cropped tight around the image. A Google account for Gemini |
Step by step
Step 1: Google: find where the image came from
In the Google app or on the Google homepage, search for the image and open the Image Viewer. Select About, then select Learn more about an image. Google's page says you may see when Google may have first seen a similar version of the image, other pages that use similar versions, and pages where it appeared much earlier. Google notes this feature is currently only available in some regions.
Tip: Google's page says to check whether an image is older or newer than you expected.
Step 2: Gemini: check for Google's AI watermark
Go to gemini.google.com and sign in. Click Add files, then Upload Files. You can check one image, video or audio file at a time, 100 MB or less. Ask: "Is this AI-generated?" Gemini looks for SynthID, an invisible watermark in content made or edited by Google's AI. Google's page says Gemini can currently recognise only content created by Google AI tools, so no watermark does not mean the image is real.
Prompt to copyIs this AI-generated?
Tip: For a screenshot, crop tight around the image. Google's page says that gives the best accuracy. The page lists a quota of about 10 image checks in any 24 hours.
Step 3: Content Credentials: read the image's history
Open contentcredentials.org/verify and drag the file into Verify. If the file carries Content Credentials, you can see how it was made and changed over time. The site says Content Credentials are still rolling out, so many files will show nothing. Gemini's check also reads them: Google's page says you can see media composition, edit history and whether AI was used.
Tip: A screenshot usually loses this information. Use the original file where you can.
Common mistakes
- Treating one result as proof. Google's page says no SynthID watermark only means the image was not made by Google's AI, and it could still come from another AI system.
- Checking a screenshot with lots of border. Crop tight around the image, as Google's page advises.
- Trusting the details shown with an image. Google's page warns that credit and digital source type details can be modified with image generation and editing tools.
Questions people ask
Can AI detectors tell for sure if an image is fake?
No. Google's own help page lists several signals, including a watermark check, reverse image search, metadata and visual clues, and says each has limits. Use more than one.
What is SynthID?
Google's help page says SynthID embeds invisible digital watermarks into content generated by Google's AI models. Gemini can check whether an image, video or audio file carries one.
What are Content Credentials?
Google's help page describes them as a digital passport that documents where content came from and how it changed. It says they are an industry standard from the Coalition for Content Provenance and Authenticity, used for both AI and non-AI content.
What should I look for with my own eyes?
Google's help page lists inconsistent or garbled text, errors in small details such as hands, teeth or background elements, inconsistent shadows and lighting, and repetitive patterns that look unnatural.
Checked against
- OfficialGemini Apps Help: Verify AI-generated images, videos, and audio
- OfficialGoogle Search Help: Learn more about an image
- OfficialContent Credentials: Verify
AI apps change their buttons often. If a step no longer matches what you see, email hello@openloopnews.com and we will update this guide.
