TL;DR
Get the latest gadgets 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 expanded public access to its SynthID Detector, making the tool available globally in English. It can check images, video and audio for SynthID watermarks in content made with AI tools from Google and named partners, but the announcement does not say it can identify all AI-generated media.
Google says its SynthID Detector is now available to people globally in English, allowing users to check whether images, video or audio carry AI watermarks associated with Google and selected partners. The expansion brings a tool previously offered in an early version to media professionals to a wider audience, though its checks cover watermarked content rather than establishing whether any media was generated by AI.
The detector checks for SynthID watermarks in images, video and audio. Google says users can use it to assess content made with AI from Google or its partners, naming OpenAI, NVIDIA and Kakao; the company says Apple support is planned soon. The announcement does not specify how the detector reports results or what users should infer when it does not find a watermark.
Google said it has watermarked more than 180 billion images and videos, as well as audio content it describes as equivalent to 240,000 years. These are company-reported cumulative totals; the announcement does not provide a measurement period or an independently verified breakdown. Google introduced SynthID in 2023 and says the technology embeds marks designed to be imperceptible.
The company also says built-in verification features in Google Search, the Gemini app and Chrome now handle more than one million requests daily. That figure refers to requests across those existing features, not necessarily use of the newly expanded detector. Google did not provide further details about the count or its measurement window.
A Wider Route to Watermark Checks
Making the detector available beyond a professional audience gives more people a way to check some online media for signs that it was made with AI. That may help readers, publishers and others seeking additional context before sharing or relying on an image, clip or recording. The tool is a verification aid, not a general test of whether content is authentic.
Its usefulness depends on media carrying a watermark the detector recognizes. Google describes checks for content made with its own AI and partner systems, but does not claim that the tool can identify every AI-generated file or every kind of manipulation. A result that detects a watermark can offer evidence of a supported tool’s involvement; a result that does not detect one cannot, on the information released, establish that content is human-made.
The global English-language release broadens access, while the language restriction means the company has not announced availability in other languages. The detector also sits alongside verification features embedded in Google products, giving users more than one route to inspect supported media, though the company has not explained how those features’ results compare.
As an affiliate, we earn on qualifying purchases.
From SynthID Launch to Public Access
Google launched SynthID in 2023 as a way to mark AI-generated media with watermarks designed to be difficult for people to see or hear. The company says it has applied the marks at large scale across images, video and audio. It introduced an early version of the detector last year, initially aimed at media professionals who need to verify content.
The latest announcement changes who can access the detector: Google says it is expanding availability to everyone worldwide in English. It names OpenAI, NVIDIA and Kakao among the partner systems whose content the detector can check, and says Apple support is expected soon. The announcement does not describe the technical arrangements with those companies or list specific supported products and file formats.
““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.””
AI-generated media verification tool
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Limits of Detection Still Unspecified
Google’s announcement does not explain the detector’s accuracy, how it handles edited or compressed files, or whether it can recognize every watermark format used by the named partners. It also does not say what a result looks like, whether users receive a confidence level, or how the tool treats files with no detectable SynthID mark.
The company has not claimed that the detector identifies all AI-generated content. The announcement gives no availability date for languages beyond English and describes Apple support as coming soon without a specific launch schedule. Details about supported file sizes, formats, access requirements and privacy handling are also absent from the material provided.
video and audio watermark detection software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Partner Support and Language Expansion
Google says Apple support is planned soon, but has not set a date. The company also has not announced when, or whether, the detector will be offered in languages other than English. Users seeking to try it can look for the SynthID Detector through Google’s announced global English-language release; the source does not give a direct access link or technical requirements.
Further information about accuracy, supported formats and the meaning of detection results would help users understand what the tool can and cannot verify. Until Google publishes those details, the detector is best understood as a way to check for certain supported watermarks, not as a definitive judgment about the origin or authenticity of any media file.
As an affiliate, we earn on qualifying purchases.
Key Questions
What does Google’s SynthID Detector check?
Google says it checks images, video and audio for SynthID watermarks associated with AI content from Google and selected partners.
Who can use the detector now?
Google says the tool is available to everyone globally in English. The announcement does not provide details about access requirements or supported file formats.
Which partner systems does Google name?
Google names OpenAI, NVIDIA and Kakao as partners whose AI-made content the detector can check. It says Apple support is planned soon, without giving a date.
Does a negative result prove that content was made by a person?
No. Google describes a check for supported watermarks; it does not say that the detector can identify all AI-generated media. The announcement does not explain what conclusions users should draw when no watermark is detected.
How much media has Google watermarked?
Google reports more than 180 billion images and videos and audio equivalent to 240,000 years. The company does not give a measurement period or independent verification for these cumulative figures.
Source: rss
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
