Is that photo real? Was this voice message actually recorded by a person? From this week, anyone can at least ask Google. The company has opened a public SynthID website that checks whether an image, a video or an audio clip was created with AI tools that use its invisible watermark.
Until now, the detector was only available to a selected circle of journalists, media professionals and researchers, who had been testing it since Google I/O last year. With the public launch, any user can upload a file and get an answer. The service handles a broad range of formats: JPG, PNG, WEBP, AVIF, HEIC, TIFF and GIF images, MP4, MOV and WEBM videos, and WAV, MP3, OGG, FLAC, AAC and M4A audio files.
SynthID, introduced in 2023, embeds a hidden signal into AI-generated content that people cannot see or hear but software can detect. Google’s own generators, including the Nano Banana image model, Veo for video and Lyria for music, as well as products such as Gemini, Flow and Vids, stamp their output with it. OpenAI, Nvidia and Kakao also support the standard, and Apple is reportedly preparing to join. Verification is already built into the Gemini app and Chrome, and Google says people run about one million checks a day.
The tool has clear limits, though. It can only recognise content carrying the SynthID watermark, so material from systems that do not use it will slip through. Microsoft and Meta rely on their own marking schemes, and experience shows that such detectors are far from infallible, sometimes failing to flag even content made by their own companies’ models.
Still, a free, open checkpoint is a useful step at a time when convincing fakes spread faster than corrections. For everyday users the best advice remains the same: treat a “not detected” result as one clue among many, not as proof that something is genuine.
