Google is now making its SynthID detector available worldwide in English to tell if media was created with AI from major providers, with Apple support on the way.
Google is now offering its SynthID AI detector to everyone in English worldwide, so you can check if media was made with AI from major providers. It looks for faint watermarks embedded in images, videos, and audio. If a watermark is detected, the detector indicates AI involvement. Apple support is expected soon.
Why it matters: it helps people gauge trust in what they see online, from photos and clips to audio recordings. It’s not perfect, some AI content may not be marked, and not all tools are covered yet, but it’s a helpful step toward clearer media provenance.
What happens next: detection for Apple-made AI media is on the horizon, and you can start using SynthID now to check media you already have.
Google has broadened access to its SynthID AI detector, letting anyone in the world check whether an image, video, or audio file was created with AI from major providers. The tool, now available in English globally, marks a big step in helping everyday readers verify the authenticity of media they encounter online.
Here’s how it works in plain terms: SynthID embeds faint, nearly invisible watermarks into media. The detector then scans for those watermarks and tells you if the file was likely AI-generated. Google says the system has watermarked over 180 billion images and videos and 240,000 years of audio content since SynthID first launched in 2023. Support for AI tools from OpenAI, Nvidia, Kakao, and Google itself is already covered; Apple’s AI tools are on the roadmap for detection as well.
For everyday users, SynthID provides a straightforward check to verify the provenance of a file you’re viewing or sharing. If a photo, video, or audio clip contains a detectable AI watermark, the detector will flag it as AI-made. That kind of transparency can help people decide how much to trust what they see, whether they’re fact-checking a news clip, evaluating a social post, or deciding whether to share content with friends and family.
It’s not a perfect shield. The detector relies on watermarks that are embedded in the media at creation time, so items that didn’t come with such marks or that have had heavy editing might not be flagged. And while Apple’s tools aren’t yet covered by the detector at launch, Google says detection for Apple-made AI media is coming soon. That keeps the door open as more AI-enabled creation options appear in the market.
At this stage, SynthID can be used on images, videos, and audio that were created with AI from Google, OpenAI, Nvidia, or Kakao. If a watermark is found, the detector confirms AI involvement; if not, it returns a neutral result. The tool is intended to be easy to use: you visit the SynthID Detector website and run a check on the file in question.
Media literacy tools like SynthID matter in the United States, United Kingdom, and Canada because misinformation and manipulated media are common online. A simple check badge can help viewers make quicker, better-informed judgments about what they’re seeing. For families, this can reduce the spread of misrepresented clips and give teachers, journalists, and health workers a clearer way to evaluate digital content.
Google notes that detection for Apple-made media is on the way, which should broaden the coverage further. In the meantime, users can start checking media they come across now, especially images and clips that seem tied to AI-driven sources. The company did not announce any price changes or new features beyond the global availability in English, so there’s no cost barrier to trying SynthID for most people.
No. Watermarks are a helpful signal, but they aren’t a perfect guarantee. Some files may lack watermarks or be altered in ways that obscure them.
It can reduce risk by making AI involvement more visible, but it doesn’t stop manipulation. Use it alongside other verification steps, like cross-checking with reputable sources.
Google says detection for Apple-created AI media is coming soon, but no exact date has been shared.
If you’re curious, try the SynthID Detector on a few items you’ve saved or downloaded. It’s a quick way to gauge whether AI played a role in their creation. For creators, consider whether adding watermarks to AI-generated media could help contextualize your work and reduce confusion among viewers.
Official information from Google about SynthID and its rollout is the basis for these updates. Details about Apple’s forthcoming detection remain unconfirmed in official statements but have been noted as forthcoming by the detector’s developers.
You're reading the quick version.
A UPS shipment carrying sensitive F-35 cockpit canopy and weapons-bay door components was allegedly rerouted through Hong Kong after an employee missed an ITAR alert, raising questions about secure handling of military tech.
Raheim Hamilton, co-creator of Empire Market, gets 40-year prison term for running a platform that facilitated hundreds of millions in illegal drug sales and other crimes.
In Kakuma, refugees use AI tools to research and translate, but pay is inconsistent and transparency is uneven. Here’s what it means for ordinary workers.
