Published : 08 Oct 2026, 11:23 PM
Updated : 08 Oct 2026, 11:23 PM
Google has launched a tool that lets anyone check whether an image, video, audio file or text was made using artificial intelligence.
The SynthID Detector works by reading imperceptible watermarks embedded in AI-generated content, reports The Independent.
Google says it has already watermarked over 180 billion images and videos and 240,000 years' worth of audio.
The tool is currently compatible with content produced on platforms from Google, OpenAI, Nvidia, Kakao and Apple, the report added.
An earlier version was made available to media professionals last year. The latest release opens access to the general public.
“As AI tools become better and easier to use, having clear context about the content you encounter online helps you make informed decisions about what to trust,” wrote Pushmeet Kohli, vice-president of science and strategic initiatives at Google DeepMind, according to The Independent.
The watermarks are invisible to human viewers, do not affect image or video quality, and are designed to survive cropping and filters, it added.
The launch comes days after OpenAI unveiled a similar watermarking system for ChatGPT text outputs, called textGrain, which subtly alters word selection so automated systems can detect AI authorship.
Anthropic has announced comparable technology for Claude, though both companies cautioned the watermarks lose effectiveness when humans edit the AI-generated text.
The rollout is partly driven by the European Union's AI Act, which requires generative AI products to make their outputs identifiable by automated systems.
The announcement coincided with the sentencing of a North Carolina man to 18 months in prison for using AI to generate hundreds of thousands of songs and fraudulently collect streaming royalties from platforms including Spotify and Apple Music over seven years.
