Google is making its powerful AI verification tech available to everyone. After keeping its detection systems locked behind enterprise paywalls and limited journalist trials for over a year, the company opened synthid.com to the public, letting anyone check whether photos, videos, or audio tracks were created or edited with generative AI.
The new web platform expands far beyond Google’s own ecosystem. While asking Gemini used to be the only way to inspect files for Google-made watermarks, the dedicated website supports invisible signatures embedded by partners like OpenAI, Nvidia, and Kakao. Apple Intelligence support is coming down the line.
Drag-and-drop checks across multiple formats
Using the site is as straightforward as dropping a file into a browser box. SynthID handles a long list of formats—including standard JPGs, high-res HEIC files, MP4 videos, and uncompressed WAV audio. Instead of forcing you to guess which company generated a clip, the portal scans for encoded signals in pixel arrangements and audio waveforms. This gives you a quick confirmation on whether synthetic tools were used.
Google also took measures to keep bad actors from using the detector to train watermark-removal scripts. The company requires users to sign in with a Google, OpenAI, or Apple account. It also caps daily scans at around ten checks per user.
The site won’t break down which exactly parts of an image were edited. However, it offers a fast binary answer without forcing you to buy enterprise software.
The ongoing battle against untracked AI content
While launching a free public checker gives everyday users a solid tool against deepfakes, watermarking isn’t a silver bullet. Competitors like Meta rely on separate detection standards, and open-weight models often generate completely untagged media.
Even so, opening up synthid.com sets a strong standard for media authenticity as synthetic media floods social feeds. By turning invisible DeepMind watermarks into a free, accessible verification tool, Google is helping users double-check questionable media before hitting share.
The post Google Unlocks SynthID to the Public so Anyone Can Spot AI Fakes appeared first on Android Headlines.
​Â
