Internet search giant Google has announced that it is making its SynthID Detector available to everyone. If you’re unfamiliar, SynthID itself is the tech used to embed invisible watermarks into AI-generated content. With that in mind, the idea is to make it easier to detect such content – just upload it to the SynthID Detector website and let it do its trick. There is a catch though, and it’s right there in the name.
Because the tool specifically checks for the invisible SynthID watermark, it only works for content that’s generated by Google’s own Gemini, or partners that also use SynthID. The list includes OpenAI, NVIDIA, South Korea’s Kakao, “and soon, Apple”, according to the announcement page. In essence, this means that if you upload a clip or image to the tool, and it tells you that “SynthID was detected”, then you can be sure that it’s definitely AI-generated to some degree, even if it was just edited using such tools.
But what if it tells you that “SynthID was not detected”? Unfortunately, that’s exactly what it says – the tool does not guarantee that the thing you uploaded was not AI-generated or edited. The page itself even tells you that “this is not a general AI detector. We can only detect media from the above companies, who have adopted SynthID technology”. Similar verbiage can be found in the FAQ section at the bottom of the page.
If you were looking for a single tool that can detect any and all types of AI generation and editing, we’re not there yet. While the partner list looks to cover pretty popular tools, there are other equally popular ones like Grok and Midjourney that are not listed. As such, media generated by them won’t be flagged by the SynthID Detector tool. Either way, you can give it a try by visiting the tool’s URL, linked below.
(Source: Google [1], [2])
