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 opened its SynthID Detector to everyone worldwide in English, extending access beyond the media professionals who could use an early version. The tool checks image, video and audio files for SynthID watermarks from Google and participating partners, but the source does not say it can identify every AI-generated file.
Google has opened its SynthID Detector to the public worldwide in English, allowing people to check whether images, videos and audio files carry AI watermarks from Google or participating partners. The expansion moves the tool beyond an earlier version intended for media professionals, but it does not establish that the detector can identify all AI-generated media.
Google says the detector can assess files made with AI from Google and partner organizations, naming OpenAI, NVIDIA and Kakao. Apple is expected to be included soon, according to Google. The company describes the system as looking for SynthID, an imperceptible watermark embedded in supported media, rather than determining authorship from the content alone.
Google reports that it has watermarked more than 180 billion images and videos, as well as 240,000 years of audio content, since SynthID launched in 2023. Those figures are company-reported totals; the announcement does not give an independent audit or explain how the volume was calculated.
The public detector joins verification features in Google Search, the Gemini app and Chrome. Google says those built-in features handle more than one million requests daily. The announcement does not specify how the new detector presents results, its file-size or format limits, or whether users can submit media in every country where the English-language tool is available.
A watermark check, not a universal test
Making the detector publicly available gives readers, publishers and other users another way to check the origin of some media they encounter online. This may help add provenance information to images, video and audio circulating without clear context, particularly when a file was created using a participating system that applies SynthID.
The scope matters: the tool checks for supported watermarks. Its results should not be read as a general verdict on whether something is AI-generated. A file without a detected SynthID watermark could be human-made, generated by a tool that does not use SynthID, or otherwise outside the detector’s coverage. The announcement does not describe how the service handles altered or partially edited files.
Google’s move also points to the value and limits of shared provenance systems. Partner participation could extend the number of tools whose media is checkable, but users still need clear explanations of what a positive or negative result means. Google has not described a common industry-wide standard or confirmed that every named partner’s output can be checked immediately.
As an affiliate, we earn on qualifying purchases.
SynthID expands beyond early access
Google introduced SynthID in 2023 as a way to mark AI-generated media with watermarks designed to be imperceptible to people. The company says the technology has been applied to images, video and audio. Its detector is a separate way for people to look for those marks after encountering a file.
Last year, Google introduced an early version of the detector for media professionals. The current announcement expands access to everyone and specifies global availability in English. It also places the detector alongside Google’s existing verification features in Search, Gemini and Chrome, which the company says receive more than one million requests per day.
The announcement names OpenAI, NVIDIA and Kakao as partners whose AI-generated content the tool can check, and says Apple will be included soon. It does not give a partner-by-partner account of which products or file types are supported, so the list should not be taken to mean every output from each company is detectable.
““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.””
As an affiliate, we earn on qualifying purchases.
Coverage and results need detail
The announcement does not explain the detector’s accuracy, how it reports uncertain results, or how it performs when media has been edited, compressed or combined with other material. It also does not list supported file formats, upload limits, data-retention practices or country-specific access restrictions.
It remains unclear which products from OpenAI, NVIDIA and Kakao are covered at launch, and what must happen before Apple support arrives. Google’s description does not claim the tool can detect AI content that lacks a compatible watermark, so a result that does not find SynthID should not be treated as proof that a file was made by a person.
As an affiliate, we earn on qualifying purchases.
Partner support and access details
Google says the detector is available globally in English starting today and that Apple support is coming soon, but it has not provided a date for that addition. Further details about supported products, file handling and how users should interpret results would help establish the tool’s practical reach.
For now, users can treat SynthID Detector as a way to check for a specific watermark in supported media. Its usefulness will depend on how broadly partners apply compatible marks and on whether Google publishes clearer guidance about positive, negative and inconclusive checks.
image and video authenticity detector
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
What does SynthID Detector check?
It checks image, video and audio files for SynthID watermarks associated with Google and participating partners. Google has not said it can identify every AI-generated file.
Who can use the detector?
Google says access is expanding from an early version for media professionals to everyone worldwide in English. The announcement does not detail country-specific restrictions.
Which companies’ media can it check?
Google names its own AI systems and partner organizations OpenAI, NVIDIA and Kakao. It says Apple will be added soon, without giving a date or listing specific supported products.
Does a negative result mean the media is not AI-generated?
No. The announcement describes checks for supported watermarks, not a universal AI detector. A file without a detected SynthID mark could still have been created using a system outside the tool’s stated coverage.
Source: rss
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
