Google opened its SynthID Detector to everyone on 7 October 2026, globally in English, it said in a blog post. Anyone can now check whether an image, video or audio file was made with AI from Google or its partners, which Google names as OpenAI, NVIDIA, Kakao and, soon, Apple.
Google says it has watermarked over 180 billion images and videos, along with 240,000 years of audio, since launching SynthID in 2023. Its built-in verification features in Search, the Gemini app and Chrome now regularly handle over 1 million requests daily, the post says.
Business Pill · A HIDDEN MARK, BUILT IN AS IT IS MADE
A one-minute explainer of watermarking: a signal added to content as it is made, invisible to people, that a checker can later read to tell where it came from. It teaches the general idea only and says nothing about any company in this story.
The key insight: As we read it, Google is turning its own watermark into a shared check. The detector now reads SynthID marks from OpenAI, NVIDIA and Kakao as well as Google, with Apple named as coming, and it is open to anyone with a Google, OpenAI or Apple login, according to Google and Ars Technica. What it can see is limited to that group: Ars notes that providers using other schemes, or none, will not be identified.
What Changed
Google says it introduced an early version of the detector last year to help media professionals verify AI-generated content. The new version is open to anyone.
Ars Technica reports that the tool lives on a new site, SynthID.com, and that until now the only way for people outside Google’s trusted tester program to check for SynthID was to ask Gemini. It also reports that the previous detector only checked for Google’s version of SynthID.
According to Ars, the public site gives a message saying whether SynthID was used to produce or edit the media, rather than highlighting the watermarked regions as the internal tool did.

The Limits Google Set
Ars reports that users must log in with a Google, OpenAI or Apple account, and quotes Google as saying each user gets “a daily usage quota of approximately 10 image, video, and audio checks.”
Ars says Google engineers told it that limiting access is necessary to prevent people from using the checks to develop SynthID removal tools.
Ars also notes that some model providers, such as Meta, use different watermarking technology and others use none, so content from those sources will not be identified by the tool. It says the website will carry a disclaimer saying as much.
The Structural Read
The scale is Google’s to state. It says it has watermarked over 180 billion images and videos and 240,000 years of audio, and that its built-in checks in Search, Gemini and Chrome handle over 1 million requests a day.
Access is rationed on purpose. Ars quotes a daily quota of about 10 checks per user and reports that Google engineers said the limit is meant to stop people using the checks to build SynthID removal tools.
The answer is deliberately coarse. Ars reports the public site says only whether SynthID was used to produce or edit the media, without the region highlighting of the internal tool.
Google, announcing the expanded SynthID Detector
“With the new SynthID Detector, anyone can easily check if an image, video, or audio file was made with AI from Google or our partners, including OpenAI, NVIDIA, Kakao, and soon, Apple.”
Three Implications
ONE CHECK, SEVERAL LABS Google says the detector reads marks from Google, OpenAI, NVIDIA, Kakao and, soon, Apple.
RATIONED TO RESIST REMOVAL TOOLS Ars reports a quota of about 10 checks a day per user, which it says Google engineers tied to preventing removal-tool development.
COVERAGE STOPS AT THE PARTNER LIST Ars notes that content from providers using other watermarking schemes, or none, will not be identified.
The Business Engineer Lens
This story maps onto the Business Engineer framework The AI Embedding Business Model.
The framework’s starting point: “Your product stops being the thing companies buy and becomes the thing other systems assume exists.”
As we read it, SynthID is moving that way: rival labs’ images now carry Google’s mark, and Google’s detector is the place to check them, so the watermark becomes a layer other AI providers build on rather than a Google feature.
What Is Not Established
We read Google’s blog post and Ars Technica’s report in full. The 180 billion and 240,000-year figures are Google’s own. Google’s post does not give an accuracy or false-positive rate for the detector, and Ars does not report one; we did not test the tool.
The quota, login and output details come from Ars, which attributes the quota wording to Google. We did not contact Google or its partners.
The Bottom Line
Google has opened its SynthID Detector to anyone, globally in English, to check images, video and audio for watermarks from Google, OpenAI, NVIDIA, Kakao and, soon, Apple. Google says it has watermarked over 180 billion images and videos, and Ars reports a daily quota of about 10 checks per user.
94,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. We read Google’s blog post of 7 October 2026 and Ars Technica’s report of the same day in full. The scale figures are Google’s; the quota, login and output details are reported by Ars. We did not test the tool or contact Google. Nothing here is a forecast, and nothing here is financial or investment advice.
Sources: Google: We’re making it easier to identify AI-generated content globally (7 Oct 2026) · Ars Technica: Google rolls out improved SynthID AI content detector, now available globally (7 Oct 2026)









