Google's SynthID Detector now checks AI watermarks from several providers
Users can check uploaded images, audio and video at SynthID.com, though a negative result does not prove content is human-made.
Google DeepMind has made its SynthID Detector generally available, letting users upload an image, audio clip or video to the SynthID.com site and check it against watermarks from partners including OpenAI, NVIDIA and Kakao, with Apple support due soon. Before this, Gemini could only flag content made by Google's own AI, and checking other providers meant using each vendor's separate tool. The result is a simple yes-or-no answer, but a negative reading does not prove that content is human-made, since some files return no confident signal. Coverage is limited to companies that have adopted SynthID, so output from Meta or Microsoft, which use different watermarking, falls outside it, and heavy edits can weaken the watermark and produce false negatives. SynthID was also named to TIME's Best Inventions of 2026 list, although commenters remain skeptical about its effectiveness.
Why it matters
Readers who need to judge whether a file is AI-generated now have one place to check watermarks from several providers instead of separate tools for each vendor. The check is a simple yes-or-no answer, and it only covers companies that have adopted SynthID. Content from providers using different watermarking falls outside it, and heavy edits can weaken the watermark, so a negative reading is not proof that content is human-made.
Matters or hype?
Does this story matter, or is it hype? Vote before you see what everyone else said.
Sources
- InfoQ