What happened
On October 7, 2026, Google opened its SynthID Detector to everyone. The portal at synthid.com is now available globally in English, letting anyone upload an image, video, or audio file and check whether it carries SynthID watermarks from Google or its partners — OpenAI, NVIDIA, and Kakao, with Apple set to follow. The announcement came via Google’s The Keyword blog, written by Pushmeet Kohli, VP of Science and Strategic Initiatives at Google DeepMind.
The details
SynthID is Google’s invisible watermarking system, launched in 2023, that embeds a hidden signal into AI-generated media at the moment of creation. Since launch, Google says it has watermarked over 180 billion images and videos, plus the equivalent of 240,000 years of audio content. The Detector is the checking half of the system: upload a file, and it tells you whether a recognized watermark is present.
An early version of the Detector was released in 2025 for media professionals only. The global release is a big step up in ambition — and in scale. The Detector now joins built-in verification features in Search, the Gemini app, and Chrome, which Google says already handle more than 1 million verification requests daily. WebProNews reports the portal accepts JPEG through HEIC images, MP4 and WEBM video, and common audio formats from WAV to M4A, with sign-in through a Google, Apple, or OpenAI account.
Google’s stated roster is the news-within-the-news: OpenAI, NVIDIA, and Kakao adopting SynthID means the detector reaches well beyond Google’s own Gemini, Imagen, and Veo outputs. Apple joining later would make it the first truly cross-industry AI watermarking check.
Why it matters
Fake AI imagery is the misinformation story of the decade — four of the top 20 most-viewed posts on Facebook in the U.S. last year were AI-generated, per TechResearchOnline. A free, one-click “was this made by AI?” check is the closest thing to a universal answer we’ve had. But read the fine print Google writes itself: the site’s FAQ states plainly, “This is not a general AI detector.” It can only identify media from companies that have adopted SynthID — a clean result doesn’t prove an image is real, just that no recognized watermark was found. Coverage from the watermark era of AI starts here, and knowing the limits is the whole point: this is a powerful tool for verifying provenance, not a lie detector.
Bottom line
Google’s SynthID Detector at synthid.com is now open to everyone in English, checking images, video, and audio against watermarks from Google, OpenAI, NVIDIA, Kakao, and soon Apple. Huge for transparency — just remember it only detects participating models, not all AI content.
Frequently asked questions
What is SynthID?
SynthID is Google’s system of imperceptible watermarks embedded into AI-generated images, video, audio, and text at creation time. The Detector reads those marks; it can’t spot unmarked AI content.
Is the SynthID Detector free?
Yes — anyone can use synthid.com globally in English. It requires signing in through a Google, Apple, or OpenAI account.
Does a clean result mean the media is real?
No. It means no recognized SynthID watermark was found. Content from models that don’t use SynthID won’t be flagged.
Which companies’ AI content can it detect?
Google’s own models (Gemini, Imagen, Veo, Lyria) plus partners OpenAI, NVIDIA, and Kakao. Apple support is coming soon, per Google.
How is this different from the earlier version?
Google released an early Detector in 2025 restricted to media professionals. Today’s launch opens it to the general public globally and adds partner watermark coverage.
Sources: Google’s The Keyword (blog.google), Unite.AI, WebProNews