Google opens SynthID AI detector globally, caps checks at ~10/day

Google's SynthID detector now puts watermark checking for its own tools, plus those of partners like OpenAI and Apple, into a single login-gated website. That convenience is real, but the design choices reveal the tension: Google engineers have told Ars that limiting access is necessary to prevent people from using the checks to develop SynthID removal tools, and the fuzziness about the exact number is likely because the system may lock you out sooner if you're checking multiple images that are very similar, which could signal someone is trying to tune a bypass workflow.

The tracker's reach is inherently partial. Some model providers use different watermarking tech, while others use no watermarks at all, and open-weight models that people can run on their own hardware produce increasingly realistic AI content that doesn't even include simple metadata labeling. That means plenty of AI content will not be identified as AI in the SynthID tool, and the website will have a disclaimer saying as much. It's also plausible that despite Google's efforts, someone will find a way to remove SynthID, making existing AI content look more legit — not a perfect system, but the reality of relying on proactive watermarking for AI images proliferating across the Internet.

An alternative path is to label real images, video, and audio at creation using cryptographic technologies like C2PA, which Google's Pixel phones are among the few devices doing. In the meantime, SynthID is the best bet you've got to identify AI fakes, though its usefulness hinges on whether the quota and the login gate keep determined bad actors from probing the system.