Google recently opened its SynthID Detector tool to everyone, allowing anyone to check whether an image, video, or audio file contains AI. The tool scans uploaded media for a hidden marker that Google embeds in AI-generated content and provides a result within seconds.
The tool not only works for AI content generated by Google, but also by OpenAI, Anthropic, Nvidia, ElevenLabs, and others. These companies use the same Synth ID to watermark their AI-generated content.
Here’s how you can use it to check content yourself.
How to Check Content for AI Watermarks
- Visit the website: Open SynthID.com to access Google’s detection tool.
- Sign in: Log in using a Google, OpenAI, or Apple account.
- Upload your file: Select the image, video, or audio file you want to check.
- Check the result: Wait a few seconds for the tool to scan the file and report whether it finds a SynthID watermark.
The result can help you determine whether the content carries Google’s AI watermark. However, a file without this watermark is not necessarily human-made, as the tool cannot detect content from every AI system.
Users are reportedly limited to around 10 checks per day.
What Can SynthID Detector Find?
Google introduced SynthID in 2023 as an invisible watermark embedded in AI-generated content. Before the detector became publicly available, ordinary users had no direct way to check whether this watermark was present.
Google first introduced a limited version of the detector for selected researchers and journalists in 2025. Its public release now gives everyone a way to check for the watermark without asking Gemini whether something was generated by AI.
The tool can also detect SynthID watermarks in AI-generated media from Google’s partners, including OpenAI, NVIDIA, and Kakao. Apple also participates in SynthID, but support for detecting Apple-generated content is still being developed.
What the Tool Cannot Detect
SynthID Detector looks for an embedded watermark rather than examining an image or video for visual signs of AI generation. This approach should reduce false positives, but it also limits what the tool can identify.
For example, it cannot detect content generated by Meta’s AI tools because Meta does not use SynthID. Content created by other AI systems that do not use the watermark may also go undetected.
The tool only reports whether it finds a watermark. It does not identify the specific parts of an uploaded file that were generated by AI.
Google says SynthID watermarks can survive edits made to AI-generated content after it is created. However, users cannot repeatedly test unlimited versions of a file, as the detector reportedly has a daily limit of around 10 checks. A Google spokesperson told Ars Technica that the restriction is intended to stop people from repeatedly testing modified files to work out how to remove the watermark.
Other Ways to Detect AI Content
SynthID is not the only system designed to identify AI-generated content.
OpenAI operates its own detector, which can identify AI-related metadata as well as SynthID watermarks. Meta uses a separate watermark called Content Seal and offers a web-based detector. However, a 2026 Reuters report found that Meta’s tool sometimes failed to identify AI-generated content correctly.
Anthropic and Microsoft have also started watermarking AI-generated media, but neither currently offers consumer-facing tools for checking those watermarks.
There is still no single detector that can reliably identify content generated by every AI system. Most available tools look for specific watermarks or metadata, while general-purpose AI detectors that analyze content itself can produce unreliable results and sometimes incorrectly label human-created material as AI-generated.
The post How to Check if Any Content is AI-Generated Using Google’s Free Tool appeared first on ProPakistani.
