Google expands AI watermark checker to everyone

The short version

Google makes SynthID Detector available to all users worldwide in English, enabling quick checks for AI-generated media.
The detector now identifies the SynthID watermark from tools including Google, OpenAI, NVIDIA and Kakao; Apple support is expected later this year.
Detector only flags media with a SynthID watermark and cannot detect all AI-created content or distinguish edits from AI creation.
This move aligns with a broader push for verifiable AI media as more people encounter AI-generated images, videos, and audio online.

Quick read · 2 min

Google is expanding its SynthID Detector to all users worldwide in English. The tool lets you upload a photo, video, or audio file to see if it carries a watermark that proves it’s AI-generated. It now covers content from Google, OpenAI, NVIDIA and Kakao, with Apple expected to join later this year.

What it means for you: if you ever question a viral clip or image, this tool offers a quick check for AI watermarks. It isn’t perfect and won’t catch every fake, but it helps separate AI-created media from real content when a SynthID watermark exists. What happens next: Apple support could broaden the ecosystem, improving coverage and usefulness for everyday consumers.

Try the detector on any English-language AI content you encounter.
Look for SynthID watermarks on media from partner tools.
Keep in mind it won’t flag content without a watermark.

Google has expanded its SynthID Detector to all users, letting people check whether an image, video or audio file contains an AI watermark. It’s live globally in English starting today, and you can try it directly through Google’s verification flow. The detector now recognizes AI content from partners beyond Google, including OpenAI, NVIDIA and Kakao, with Apple expected to join the list later this year.

SynthID was introduced in 2023 as a way to embed an imperceptible watermark into AI-generated media. The detector reads that watermark so a viewer can confirm the media was created by an AI tool. Google says it has watermarked more than 180 billion images and videos and 240,000 years of audio across its efforts so far, underscoring how much AI content is already out there.

Previously, the detector was aimed at media professionals who needed to verify content. Now the feature is available to anyone, and it sits alongside Google’s other verification tools in Search, the Gemini app and Chrome. Google notes these tools now handle more than one million requests a day, showing clear demand for a quick way to check authenticity.

There is a catch, though. The detector can only flag content that carries a SynthID watermark. If a fake image or clip is created with a tool that doesn’t use SynthID, the detector won’t flag it. In other words, this isn’t a catchall for every fake, but it’s a meaningful step when used with supported tools and partners.

For everyday readers, what this means is you have a faster way to check whether media you encounter online has a recognized AI watermark. If you see a watermark, the detector can confirm it’s AI-generated. If there’s no watermark, it doesn’t prove the media wasn’t AI-assisted, but it’s a signal you should be more skeptical and look for other cues.

As more AI tools appear and get integrated into our feeds, support from other big players makes these checks more practical. Apple’s impending support would extend this approach into devices and apps many households use daily.

What happens next is largely about adoption. Apple’s confirmation and the ongoing integration into more platforms will influence how widely average users rely on SynthID checks in the coming months.