In an era of exponential growth in generative artificial intelligence, the ability to distinguish between original and AI-created content has become crucial. Google has responded to this challenge by launching a new web platform that allows the general public to verify the origin of images, videos, and audio recordings. After a testing phase with selected journalists, media professionals, and researchers, this tool is now fully available to all users, representing a significant step towards greater transparency in the digital environment.
This verification tool supports a wide range of media formats. For images, it includes standards such as JPG, JPEG, PNG, BMP, WEBP, AVIF, HEIC, HEIF, TIFF, TIF, and GIF. For videos, it is compatible with MP4, MOV, and WEBM. Audio files can be verified in WAV, MP3, OGG, FLAC, AAC, and M4A formats. This broad support ensures that the tool is practically usable for most common digital media.
Haben Sie eine Idee für eine Website oder App?
Lassen Sie uns unverbindlich darüber sprechen – ohne Druck, mit einem konkreten nächsten Schritt.
The foundation of the entire system is SynthID technology, which Google introduced back in 2023. SynthID functions as a digital watermark that is invisibly embedded into generated media without affecting their quality. All of Google's generative models and tools, such as Nano Banana, Veo, Lyria, as well as tools with generative capabilities like Gemini, Flow, ProducerAI, and Vids, automatically apply this watermark to all content they create. Furthermore, SynthID support is expanding among other technology giants, including OpenAI, Nvidia, and Kakao, with early implementation also expected at Apple. However, it should be noted that OpenAI operates its own independent content checking platform.
In addition to the specialized website, the SynthID verification feature is integrated directly into the Gemini app and Google Chrome browser. The popularity of this tool is considerable, evidenced by the fact that it records approximately one million content verification requests daily. This underscores the growing public need and interest in transparency in the digital space.
It is important to realize that while efforts to verify content origin are commendable, limitations still exist. Even though companies like Microsoft and Meta are developing their own standards for watermarking and verification, these tools are not always 100% reliable. Sometimes they fail to identify content created by their own models. This indicates that the technology for recognizing AI-generated content is constantly evolving and is still subject to research and improvement.