Google opens AI content detector to users worldwide

Google has expanded access to its SynthID Detector, allowing users worldwide to check images, videos and audio files for signs of AI-generated content as concerns mount over the spread of synthetic media online.

The tool is now available globally in English and can help identify content created using supported AI systems from Google and partners, including OpenAI, NVIDIA and Kakao. Apple is expected to join the initiative, according to Google.

Google DeepMind Vice-President of Science and Strategic Initiatives, Pushmeet Kohli, announced the expansion in a blog post published on October 7, 2026.

Kohli said the move was aimed at helping users make informed decisions about the content they encounter as AI tools become more sophisticated and accessible.

Google introduced SynthID in 2023 to identify AI-generated content through imperceptible digital watermarks embedded in images, videos and audio.

According to Kohli, the technology has been used to watermark more than 180 billion images and videos, alongside audio content equivalent to 240,000 years.

The latest expansion builds on an earlier version of the detector introduced last year to help media professionals verify AI-generated material.

Users can submit supported media files through the SynthID Detector to check for signs that they were created using compatible AI systems.

Google said the tool complements verification features already available in its Search service, Gemini application and Chrome browser, which collectively handle more than one million requests daily.

The company said the initiative would provide users with additional context about digital content and help them assess what they encounter online.

The expansion comes amid growing concerns about the use of AI-generated images, videos and audio to spread misleading information, impersonate individuals and distort public understanding of events.

However, the detector’s findings should not be treated as definitive proof that content is authentic or fake, particularly when media may not carry the watermarks the technology is designed to detect.