Google Opens SynthID Detector Globally for AI Content
Google has opened its SynthID Detector to users worldwide, expanding access to a tool designed to identify AI-generated images, videos, and audio. The company announced the wider rollout on October 7, 2026, after previously testing the detector with journalists and media professionals.
The detector is now available globally in English. In addition, it can check content carrying SynthID watermarks from Google and several partner companies. The supported partners include OpenAI, NVIDIA, and Kakao, while Apple support is expected soon.
SynthID Expands Beyond Google AI
Google introduced SynthID in 2023 as a way to embed invisible digital watermarks into AI-generated content. Since then, the technology has expanded across images, video, audio, and text.
Importantly, the watermark does not appear to viewers. Instead, SynthID places an imperceptible signal within the content itself, allowing compatible detection systems to identify it later. For images and video, the system is designed to keep the watermark detectable after common edits such as cropping, filtering, and compression.
Google says it has now watermarked more than 180 billion images and videos. Meanwhile, the company says SynthID covers 240,000 years of audio content.
The wider partner network also marks a shift in how the technology can be used. Previously, SynthID detection focused largely on content produced through Google’s own AI systems. Now, the public detector can also identify compatible watermarks from other AI providers.
However, the tool does not function as a universal AI detector. It can only identify content that contains a compatible SynthID watermark. Therefore, media generated by systems that do not use SynthID may not receive a positive result.
A Wider Push for AI Content Verification
Google is also connecting SynthID with verification features across its broader product ecosystem. The company says built-in verification tools in Search, Gemini, and Chrome now handle more than one million requests each day.
Furthermore, Google has been expanding SynthID through industry partnerships. In May 2026, the company said OpenAI, Kakao, and ElevenLabs were bringing SynthID technology to more AI-generated content. Google also introduced an AI Content Detection API through Google Cloud to help organisations identify AI-generated media across their own platforms.
The partnership with NVIDIA is another part of that expansion. NVIDIA has used SynthID to watermark AI-generated video from its Cosmos world foundation models. Meanwhile, Apple is expected to join the detector’s supported ecosystem in the near future.
Consequently, Google is positioning SynthID as part of a broader content-provenance effort rather than as a standalone detection system. The company is also participating in the Coalition for Content Provenance and Authenticity, or C2PA, which develops standards for tracking how digital media is created and modified.
What the Global Detector Means
The public rollout gives users a simpler way to investigate suspicious digital media. Instead of relying only on platform-specific verification tools, users can submit supported images, video, or audio to a single SynthID detection service.
Still, a negative result does not prove that content was created by a person. The file may come from an AI model that does not use SynthID, or its watermark may have been weakened through extensive manipulation. Google therefore presents watermark detection as one part of a wider approach to understanding digital content.
The expansion comes as generative AI makes synthetic media increasingly difficult to distinguish from conventional photography, recordings, and video. Therefore, provenance technologies such as SynthID could become increasingly important for news organisations, researchers, platforms, and everyday users.
For Google, the latest rollout represents a broader effort to make AI-generated content easier to identify and to encourage more AI companies to adopt interoperable transparency technologies. As more providers adopt SynthID, the detector’s usefulness could extend beyond Google’s own AI ecosystem.
