Determining whether a piece of content was produced by a human or by AI is increasingly difficult. Google has expanded access to its SynthID web tool, which attempts to identify images, video and audio that were generated by AI, by opening the service to the general public.
Previously, SynthID had been available to journalists, media professionals and researchers, but access for the wider public was restricted. The newly public interface allows anyone to upload a file for inspection by clicking the "Select file" button in the bottom-right of the main menu and choosing the image, video or audio to be checked.
Supported file formats include: JPG, JPEG, PNG, BMP, WEBP, AVIF, HEIC, HEIF, TIFF, TIF, MP4, MOV, WEBM, WAV, MP3, OGG, FLAC, AAC, M4A.
According to Google, SynthID can detect content generated by Google, Nvidia and Kakao AI models. In practice, using the service requires signing in with a Google, Apple or ChatGPT account. The interface is currently available only in English.
Making SynthID publicly reachable is a notable step toward wider access to tools for spotting synthetic media. It lets users across sectors perform quick checks on questionable content, which can be particularly useful in journalism and research. At the same time, the requirement to log in and the single-language interface limit accessibility for some potential users.
Why this matters
- Public availability increases access to verification tools and can speed up the process of checking suspicious media.
- Broad format support (images, video, audio) means different media types can be assessed on a single platform.
- The sign-in requirement and English-only interface remain practical barriers for users without the specified accounts or sufficient language skills.
The public launch of SynthID illustrates how tech companies are releasing instruments to identify AI-generated content, while highlighting that practical limitations—access and language—still shape who can benefit from such tools.
