Google has expanded public access to its SynthID Detector, giving people around the world a new way to check whether images, videos, and audio contain digital watermarks associated with AI-generated content. The rollout, announced on October 7, 2026, marks a broader push to help users verify the origins of digital media as synthetic content becomes increasingly difficult to distinguish from authentic recordings.
Previously available in an early version to journalists and other media professionals, the tool is now accessible globally in English. Google says it can identify SynthID watermarks in content generated by its own AI systems and supported partner platforms.
A watermark-based approach to AI verification
SynthID, launched by Google DeepMind in 2023, embeds an imperceptible digital signal into AI-generated media. Unlike visible labels, these watermarks are designed to remain undetectable to ordinary viewers while allowing compatible tools to recognise them.
For images and videos, the signal is incorporated into the visual content, while audio files carry it within their sound waveforms. Google says the watermarks are designed to remain detectable after common alterations, including cropping, compression, applying filters, and changing video frame rates.
With the expanded detector, users can upload supported media files to check for a SynthID watermark. The result can provide additional context about a file's origin, helping people assess content they encounter online.
Google says it has embedded SynthID watermarks in more than 180 billion images and videos, as well as the equivalent of 240,000 years of audio, since the technology was introduced.
Support extends to AI developed by other companies
The latest expansion also broadens SynthID Detector's reach beyond Google's own AI products.
According to Google, the tool supports watermarks associated with content generated by partner companies, including OpenAI, NVIDIA and Kakao. Apple is expected to join the initiative in the future.
The wider compatibility reflects a growing need for verification methods that can work across different AI platforms. As generated video, images and audio become more realistic, identifying a watermark could give users another way to investigate whether a piece of media originated from a supported AI system.
However, the detector's usefulness depends on the content carrying a recognised SynthID watermark. It should not be confused with a universal tool capable of identifying every image, video or audio recording produced using artificial intelligence.
Google expands verification across its products
The public launch builds on Google's broader effort to make AI-content verification available through products people already use.
The company has incorporated verification features into Search, the Gemini app and Chrome. Google says these built-in features now handle more than one million requests each day, indicating growing demand for information about the origins of online media.
The strategy focuses on establishing a traceable signal when supported AI content is created, rather than relying solely on software to examine a finished file and guess whether it was generated artificially.
The limits of AI-content detection remain
Despite the wider rollout, SynthID Detector cannot establish the origin of every piece of digital media. A file without a detectable SynthID watermark may be human-created, generated by an unsupported AI system, or produced using a tool that does not embed the watermark.
The absence of a watermark, therefore, is not proof that content is authentic. Similarly, detecting a watermark provides evidence of a connection to supported AI-generated content, but users may still need additional context to understand how a file was created or subsequently edited.
For news organisations, researchers and everyday internet users, the tool offers another way to investigate questionable media. Its value will depend partly on how widely watermarking is adopted across the AI industry and how consistently verification tools can identify supported signals.
Google's expansion of SynthID Detector is a step towards making AI provenance easier to check, but it does not resolve the wider challenge of online misinformation. Reliable verification will continue to require a combination of technical tools, source checks and careful judgement about the content people encounter.
Source: Google