Fact-Checked by Irfan Ahmad
Google is expanding access to its SynthID Detector globally, with the tool available in English to help users check if an image, video or audio file was made with AI from Google or its partners, according to a Google blog post published Oct. 7.
Google introduced an early version of the detector last year to help media professionals verify AI-generated content. The company said its SynthID technology uses imperceptible watermarks across images, video and audio to help people identify AI-generated media.
Since launching SynthID in 2023, Google said it has watermarked more than 180 billion images and videos, along with 240,000 years of audio content.
The detector can check content made with AI from Google and partners including OpenAI, NVIDIA and Kakao, with Apple coming soon, Google said.
The tool joins Google’s built-in verification features in Search, the Gemini app and Chrome browser, which regularly handle more than 1 million requests daily.
Screenshot: DIW. CC BY
In a DIW test, the detector produced different results depending on how the media was submitted. When an image containing Google AI-generated or edited content was uploaded as a standalone image, the detector reported that SynthID was detected and said the media was made or edited with Google AI. However, when a screenshot (such as the one featured in this article), containing a decent part of that same content was tested, SynthID was not detected. This was also the case when the original AI-generated image, which had been identified by SynthID, was placed inside a new image covering more than 50% of its area. The detector said it was unlikely that the media was made with AI from a SynthID partner, while noting that the media could have been generated by a non-partner organization or before a partner adopted SynthID.
Screenshot: DIW. CC BY
Editor’s Note: Post updated with an additional screenshot and DIW test results.
Read next:
• LMU Study Finds People See AI as Aware, But Not Truly Conscious
• Using AI for Health Advice? Know What It Can and Can’t Do
Google is expanding access to its SynthID Detector globally, with the tool available in English to help users check if an image, video or audio file was made with AI from Google or its partners, according to a Google blog post published Oct. 7.
Google introduced an early version of the detector last year to help media professionals verify AI-generated content. The company said its SynthID technology uses imperceptible watermarks across images, video and audio to help people identify AI-generated media.
Since launching SynthID in 2023, Google said it has watermarked more than 180 billion images and videos, along with 240,000 years of audio content.
The detector can check content made with AI from Google and partners including OpenAI, NVIDIA and Kakao, with Apple coming soon, Google said.
The tool joins Google’s built-in verification features in Search, the Gemini app and Chrome browser, which regularly handle more than 1 million requests daily.
In a DIW test, the detector produced different results depending on how the media was submitted. When an image containing Google AI-generated or edited content was uploaded as a standalone image, the detector reported that SynthID was detected and said the media was made or edited with Google AI. However, when a screenshot (such as the one featured in this article), containing a decent part of that same content was tested, SynthID was not detected. This was also the case when the original AI-generated image, which had been identified by SynthID, was placed inside a new image covering more than 50% of its area. The detector said it was unlikely that the media was made with AI from a SynthID partner, while noting that the media could have been generated by a non-partner organization or before a partner adopted SynthID.
Editor’s Note: Post updated with an additional screenshot and DIW test results.
Read next:
• LMU Study Finds People See AI as Aware, But Not Truly Conscious
• Using AI for Health Advice? Know What It Can and Can’t Do
