Free tools Windows power users keep installed
One-click scans. No signup required.
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Since October 7, 2026, anyone in the world can use Google’s SynthID Detector, in English, to check whether an image, video, or audio file carries a supported AI watermark. The result is a provenance signal: it tells you whether a file was made or edited with a watermarking AI system Google recognizes. It does not tell you whether the file is authentic, whether a claim in it is true, or whether an AI system without a supported watermark was involved.
What Google changed on October 7, 2026
Google DeepMind had been testing the SynthID Detector with journalists and media professionals. On October 7, 2026, it announced that the tool was expanding to everyone globally, in English. Users can check whether an image, video, or audio file was made with AI from Google or from named partners OpenAI, NVIDIA, and Kakao. The announcement describes Apple support as coming soon. Google also says its verification features in Search, Gemini, and Chrome together handle more than one million requests a day.
Google’s announcement cites two scale figures for its SynthID work: over 180 billion images and videos, and 240,000 years of audio content. The text does not define exactly what those totals count, so treat them as Google’s own figures rather than independently measured numbers.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteDeepMind’s Pushmeet Kohli, VP, Science and Strategic Initiatives, framed the launch this way: “As AI tools become better and easier to use, having clear context about the content you encounter online helps me make informed decisions about what to trust.”
#1 Best Overall
A broader set of tools came earlier. Google’s May 19, 2026 post described SynthID verification in Gemini, expansion of verification into Search and Chrome, C2PA Content Credentials, and a new Cloud AI Content Detection API for organizations. That post also reported a cumulative figure of 50 million uses of Gemini image, video, and audio verification. That is a running total, separate from the daily request count above.
These pieces belong to one verification effort, but they are not one tool. Each checks something different, and each has its own access rules.
Four checks that are easy to confuse
The table below separates the four checks by what they look at, what they accept, who can use them, and what a result means.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Fix the driver behind crashes, sound loss and screen glitches3Clear out junk files and repair common Windows errors| Check | What it looks for | Media accepted | Access | What a result means |
|---|---|---|---|---|
| SynthID Detector (standalone portal) | Supported SynthID watermarks from Google, OpenAI, NVIDIA, and Kakao; Apple described as coming soon | Images, video, and audio | Open to everyone globally, in English, from October 7, 2026 | Supported watermark found or not found. Upload size limits, usage quotas, and retention terms: not stated in the October 7, 2026 announcement |
| Gemini verification | Watermarks associated with Google AI models only. Google says Gemini currently recognizes content created by Google AI tools, even though other companies have started adopting SynthID | Images, video, and audio, within the limits listed below | Gemini Apps; signed-in personal or qualified Workspace account | Detected: all or part of the file may have been created or edited by Google AI. Not detected: Google AI watermarking was not found; other AI systems are not ruled out |
| Content Credentials (C2PA) | A C2PA record of origin and edit history, including whether AI was involved, when compatible credentials are present | AI and non-AI media, if credentials are attached | Metadata carried inside the file; not a Google detection method | Provenance context. Credentials do not by themselves establish that a file is AI-generated |
| Cloud AI Content Detection API | Pixel-level artifacts, noise patterns, and spectral anomalies. It does not check C2PA metadata | JPEG, PNG, and WebP images | Private Preview for organizations | A probabilistic estimate that can include false positives and false negatives |
Reading the result correctly
Most mistakes come from reading a result as more than it is. The three outcomes below need different interpretations.
A positive result
A positive result means the file carries a watermark from a supported system. Google’s Gemini guidance says a detection means all or part of the file may have been created or edited by Google AI, so a positive result can describe partial editing rather than a fully generated file. It is still only a provenance signal. It says nothing about whether the content is accurate, in context, or of good quality.
A negative result
A negative result does not prove that a file is human-made. The watermark may be absent, may belong to a system the tool does not support, or may have been weakened. Google says watermarks usually survive resizing, recoloring, and compression, but repeated alterations can make them undetectable. Simple or abstract images may not contain enough detail to hold a watermark, and a minor edit may be too small to leave a detectable signal.
Content Credentials
Content Credentials are a separate mechanism. They are signed metadata that can record where a file came from and how it was edited, including whether AI was involved, but only when the file still carries them. Stripping metadata is easy, so an absent credential tells you little, and a present one needs to be checked against its source. Google’s position is that credentials add provenance context but do not by themselves show that a file is AI-generated.
Using Gemini’s verification workflow
Google’s Gemini Apps help page sets out requirements and limits for Gemini checks. They apply to Gemini, and the announcement does not say that they apply to the standalone portal.
- Account: you must be signed in with a personal Google account or a qualified Workspace account.
- Files: one file at a time, 100 MB or smaller.
- Video: under 90 seconds.
- Audio: under one hour.
- Approximate quotas in a rolling 24-hour period: 10 image checks; 10 video checks, up to 5 minutes in total; 10 audio checks, up to 3 hours in total.
The help page suggests three ways to phrase the question: “Was this image/video/audio created or edited by Google AI?”, “Is this AI-generated?”, or “Is this a real video?” The first is the most accurate fit for a watermark check. The second goes beyond what a watermark can establish, and the third asks about truth or authenticity, which a provenance check does not answer.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.The Cloud AI Content Detection API
The Cloud API is a different kind of tool. It does not look for a watermark. It analyzes an image’s pixel-level artifacts, noise patterns, and spectral anomalies, then returns a probabilistic estimate. Its current documentation, last updated October 7, 2026, lists these facts:
- Status: Private Preview.
- Accepted formats: JPEG, PNG, and WebP.
- Metadata: it does not detect C2PA metadata.
- Retention: according to the current documentation, it does not retain processed images.
- Reliability: estimates can include false positives and false negatives.
Google says the API should be supplementary evidence. It should not be the sole basis for takedowns, sanctions, or other critical actions.
Combining the checks for an important decision
- Confirm the file type and size fit the tool you plan to use. For Gemini, that means the limits above.
- Run the SynthID check and record the exact result, including whether it came from the standalone portal or Gemini, since the two may cover different systems.
- Look for Content Credentials and read what they say about origin and edits. Treat missing credentials as neutral.
- Trace the file back to its source and context, such as the original publisher, the upload date, and whether other copies exist.
- Only for images in supported formats, and only with an organizational setup, consider the Cloud API as one more probabilistic input.
No single check settles whether a file is real, altered, or accurate. The result that matters is the combination of signals, read against what each one can and cannot tell you.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
