TL;DR
Get monitors, keyboards and dev gear delivered free — and shop member deals
- Fast, free delivery on millions of items
- Access to Prime Big Deal Days deals on October 6–7
- Prime Video, Amazon Music and more included
Google has opened its SynthID Detector to people globally in English, expanding a tool previously available in an early version to media professionals. It can identify SynthID watermarks in images, video and audio made with AI from Google and named partners, but the announcement does not say it can detect every AI-generated file.
Google has opened its SynthID Detector to anyone worldwide in English, expanding a tool that was previously offered in an early version to media professionals. The detector checks images, video and audio for watermarks associated with AI content from Google and participating partners, giving users another way to assess where online media came from.
Google says users can submit an image, video or audio file to check whether it was made with AI from Google or its partners. The company names OpenAI, NVIDIA and Kakao among the partners whose content the tool can check, and says support for Apple is coming soon. The announcement does not specify when Apple support will be added.
SynthID uses imperceptible watermarks embedded in media, according to Google. The detector looks for those marks; it is not described as a general-purpose test that can establish whether any file was created or altered by AI. Google’s announcement also does not detail which file formats, maximum file sizes or editing conditions the public tool supports.
Google reports that it has watermarked more than 180 billion images and videos and 240,000 years of audio content since SynthID launched in 2023. Those are company-reported totals, not independent measurements of how much AI media exists online. Google also says verification features in Search, Gemini and Chrome handle more than 1 million requests daily; it did not provide a separate breakdown for each service.
A Check for Watermarked Media
The global English-language release makes a media-authenticity check available beyond professional users, potentially helping journalists, researchers and ordinary viewers investigate the origin of some files. Wider access could make watermark checks easier to use when AI-generated media circulates without clear labeling.
The scope is also important: a result depends on whether a file carries a watermark the detector recognizes. Google’s description names its own and partner systems, so the tool should not be treated as a definitive verdict on all content. A file without a detected mark may still have been made or edited with AI, while the announcement does not explain how results account for cropping, compression or other changes.
As an affiliate, we earn on qualifying purchases.
SynthID Since Its 2023 Launch
Google launched SynthID in 2023 as a way to mark AI-generated media with watermarks intended to be imperceptible to people. The company says the marks can be used across images, video and audio. The detector is a related verification tool: Google introduced an early version for media professionals last year, before announcing public global access in English.
The detector joins verification features Google says are already built into Search, Gemini and Chrome. The company’s statement presents these tools as part of a broader effort to provide context about media encountered online. It does not describe the specific verification experience in each product or explain whether those features rely on the same detection methods.
““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.””
media authenticity verification tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Coverage and Detection Limits
Google has not published details in the announcement about the detector’s accuracy, error rates or coverage limits, nor does it say how it handles media that has been modified after creation. It is also unclear whether the tool can identify watermarks from every version of the named partners’ systems or how users will be shown uncertain results.
The announcement does not say whether a negative check means content is non-AI-generated; it describes detection of watermarks, not universal identification of AI content. The public release is in English, and Google has not announced other interface languages. Apple support is described as forthcoming, with no release date given.
AI-generated media detection software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Partner Support and Language Expansion
The immediate next development identified by Google is adding Apple support to the detector; the company has not provided a timetable. The tool is available globally in English, while any expansion to additional languages or media sources remains unannounced.
Users and media organizations can begin trying the public detector, but its practical reach will depend on which files carry supported watermarks and how clearly the tool reports its findings. Google has not announced a schedule for publishing performance data or further technical details, so those points remain open.
video and audio watermark detection
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Who can use Google’s SynthID Detector?
Google says the tool is now available to everyone globally in English. It was previously introduced in an early version for media professionals.
What kinds of files can it check?
Google says it can check images, video and audio for SynthID watermarks associated with AI content from Google and participating partners.
Does a negative result prove that content is not AI-generated?
No such guarantee is stated in Google’s announcement. The detector checks for watermarks it recognizes, and the company does not describe it as a test for all AI-generated content.
Which partners are included?
Google names OpenAI, NVIDIA and Kakao. It says Apple support is coming soon but does not give a date.
How widely has Google used SynthID?
Google reports watermarking more than 180 billion images and videos and 240,000 years of audio since the system launched in 2023. These are company-reported figures.
Source: rss
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
