How do I use SynthID Detector?

You access the SynthID Detector and log in with Google, Apple, or ChatGPT. Then you upload a file for detection.

If SynthID is not detected, does that mean the content is human-made?

No. SynthID Detector only finds the SynthID watermark, so content from AI tools that do not embed SynthID cannot be identified.

What does the detector show when it finds a watermark?

In testing, an image made with Google's Gemini was flagged as detected and correctly attributed to Google's AI. A ChatGPT-made image was labeled as made or edited with OpenAI.

Image GenerationVideo GenerationAI Safety & AlignmentGIGAZINE AI3m agoOct 8, 2026, 16:00 JST

Google opens SynthID Detector worldwide

Send on LINE

3 Key Points

What happened
Google said on October 7, 2026 that SynthID Detector is now available worldwide in English, after limited testing. Users log in with Google, Apple, or ChatGPT to upload a file for detection.
Why it matters
The tool gives a way to check whether uploaded content carries a SynthID watermark, rather than relying on guesswork. A watermark not showing up does not prove content is human-made.

WHO IT HITSPeople who handle images and videos as part of their work, such as editors, publishers, and social media teams, may get an easy way to check whether a file from a text-to-image tool carries Google's watermark. The check only covers SynthID-supporting models, so a negative result cannot be read as proof of human authorship.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, in your inbox every morning.

Get news like this every morning, free

Context & Analysis

Google says that since it released SynthID in 2023 it has watermarked 180 billion images and videos, and 240,000 years of audio content, with 1 million daily verification requests. That scale is the backdrop for opening the detector to the public.

In a test, an image generated with Gemini prompted detection and a correct attribution to Google's AI. An image made with the same prompt in ChatGPT was labeled as made or edited with OpenAI. After cropping to just the cat and editing brightness, SynthID was still detected, but the tool no longer attributed it to OpenAI.

A comic panel from GIGAZINE's series produced no SynthID detection. The detector only looks for SynthID, so content from generative AI that does not embed it cannot be identified, and a miss does not prove the content is human-made.

FAQ

How do I use SynthID Detector?: You access the SynthID Detector and log in with Google, Apple, or ChatGPT. Then you upload a file for detection.
If SynthID is not detected, does that mean the content is human-made?: No. SynthID Detector only finds the SynthID watermark, so content from AI tools that do not embed SynthID cannot be identified.
What does the detector show when it finds a watermark?: In testing, an image made with Google's Gemini was flagged as detected and correctly attributed to Google's AI. A ChatGPT-made image was labeled as made or edited with OpenAI.

GIGAZINE AIRead Original Article

Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Get it free

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.