Google opens SynthID Detector website to identify AI-generated media
In short
Google has opened its SynthID Detector website to everyone, letting users check whether images, video or audio were created or edited with AI. The free tool recognises SynthID watermarks from Google, OpenAI, Nvidia and Kakao tools, and Apple support is expected soon. The outlets note the detector cannot identify all AI content, since watermarking standards differ and watermarks can be stripped.
Read the full story 2 min read
Google has opened its SynthID Detector website to everyone, allowing users to check whether an image, video or audio clip was created or edited with AI, according to Engadget and TechCrunch. Engadget reported the tool is free but requires users to log in with a Google, OpenAI or Apple account. TechCrunch listed the supported formats, including JPG, PNG and HEIC images, MP4 and MOV video, and WAV, MP3 and AAC audio. Ars Technica summarised the rollout as an improved detector that is now available globally. [ 1 , 2 , 3 ]
The site flags content created with AI tools from Google, OpenAI, Nvidia and Kakao, Engadget reported. TechCrunch said Google uses SynthID, introduced in 2023, to watermark media generated by its Nano Banana, Veo and Lyria models and by tools such as Gemini, Flow, ProducerAI and Vids. OpenAI, Nvidia and Kakao also support SynthID, and OpenAI maintains its own site for checking content, TechCrunch reported. Both Engadget and TechCrunch said Apple support is coming soon. [ 1 , 2 ]
Google said people currently make 1 million requests a day to verify content, according to TechCrunch, and the company has baked SynthID verification into the Gemini app and Google Chrome. Engadget reported that Google has also offered detection through its Circle to Search feature and the Gemini app, and that the dedicated website makes the capability easier to reach for people who do not already use Google's AI tools. [ 1 , 2 ]
Access was previously restricted. TechCrunch reported the tool was made available to selected journalists, media professionals and researchers for testing at Google I/O last year, and Engadget said Google had been working on SynthID Detector for more than a year. [ 1 , 2 ]
The outlets noted limits. Engadget said the detector cannot identify all AI-generated content because companies rely on different watermarking standards, that watermarks can be stripped in some situations, and that the site does not distinguish between content entirely created with AI and content merely edited with it. TechCrunch said Microsoft and Meta use their own standards and that such tools are not infallible, often failing to identify content made by their own makers' models. Engadget concluded that verifying AI-generated content is still far from foolproof. [ 1 , 2 ]
Why it matters
A public detection site makes watermark checking easier for people who do not already use Google's AI tools, Engadget reported. But because detection relies on a patchwork of company standards and watermarks can be removed, the outlets say verifying AI-generated content remains far from foolproof.
Key facts
- Google launched a website, SynthID Detector, that anyone can use to check whether media was created with AI. [ 1 , 2 , 3 ]
- The tool can flag images, video and audio generated with AI tools from Google, OpenAI, Nvidia and Kakao, according to Engadget and TechCrunch. [ 1 , 2 ]
- It is free to use but requires logging in with a Google, OpenAI or Apple account, Engadget reported. [ 1 ]
- Google says support for images created or edited with Apple Intelligence is coming soon, according to Engadget and TechCrunch. [ 1 , 2 ]
- Google said people make 1 million requests a day to verify content, TechCrunch reported. [ 2 ]
- Engadget reported the tool cannot detect all AI-generated content and does not distinguish between content entirely created by AI and content merely edited with it. [ 1 ]
Confirmed by several sources
Still unclear
- Whether Apple Intelligence image support has launched Engadget says support is coming “soon” and TechCrunch says Apple is “said to be adding support soon,” so neither outlet states it is available now.
- How the site handles content that is AI-edited rather than fully AI-created Only Engadget states the site does not distinguish between the two, so this is a single-source claim.
- The exact launch day TechCrunch says Google launched the site “on Tuesday,” while both TechCrunch and Engadget published their reports on Wednesday, 7 October.