Last updated: 12 November 2026

Free tool: runs in your browser

AI watermark and label decoderwhat a SynthID, C2PA or platform label actually means

TimeBinder adds a layer of protection against AI and human fraud and disputes. Translate a watermark result or an AI label into what it can, and cannot, show.

Nothing you choose is sent anywhere. This page does not read your file and it cannot detect SynthID.

Explain the result you have

Pick the closest match to the result you already have. You get a shareable explanation and the limits, stated plainly.

Your choices stay on this page. Nothing you tick or select is sent anywhere or tied to your name.

What result do you have?

Pick the closest match. This page explains what that kind of result can and cannot show. It does not read your file and it cannot detect a SynthID watermark. For a SynthID check, use Google's official detector, linked below.

What this does, and what it does not

What does a SynthID watermark or an AI label actually prove? The short answer

A found SynthID watermark is a signal that a participating generator marked the file. A C2PA manifest declares a recorded history. A platform label is a disclosure. None of them, on their own, prove that the contents are true, or who owns the work.

A missing mark proves even less: most files, real and generated, still travel with no watermark and no label. If the file is yours, a dated record of the original is a different kind of evidence from any of these signals.

Why this came up now

On 15 October 2026, TIME named Google DeepMind's SynthID one of its Best Inventions of 2026. Google said the system had watermarked more than 180 billion images and videos, and it points people to an official detector. The same month, DeepMind showed photoreal interactive worlds (Genie 4) that it says carry SynthID marks.

The honour made the tool more visible. It did not make watermarks universal, and it did not make a missing mark a test of authenticity. This decoder exists so a result you already have is not over-read.

Sources: Google's 15 October announcement, TIME's Best Inventions entry, the DeepMind SynthID page, and DeepMind's post on X.

What each signal does and does not show

Use the official tools for the marks they were built to read. This page does not replace them.

Watermarks, manifests and labels compared
Signal	What it can show	What it cannot show
Official SynthID Detector: mark found	That a participating system marked the file or a descendant of it.	Who made it, that the scene is true, or that every generator uses SynthID.
Official SynthID Detector: no mark	That this detector did not find a mark it can read.	That a person made the file, or that the file is genuine.
Content Credentials (C2PA)	A declared history, if the signature verifies in an official tool.	The truth of the scene, or anything about a file whose manifest was stripped.
Platform AI label	That the platform or the uploader flagged generative AI.	Authorship, ownership, or a consistent result on every app.
Unofficial "AI detector" score	Very little you should act on.	Whether a file is real or generated.
A TimeBind of your original	That an exact file existed no later than a recorded time and still matches.	Whether AI was used, who made the file, or who owns it.

Official places to check a mark

SynthID: start from DeepMind's SynthID page, which describes the official detector. This TimeBinder page cannot run that check.
Content Credentials: use our in-browser checker to read declared fields, then the official Verify tool to check the signature.

This is general information, not legal advice.

If the file is one you made

Marks and labels can be added, stripped or never applied. The file you can still show later is the original you kept:

Freeze the camera original, the export, or the mix you care about.
Make a TimeBind of that exact file. The fingerprint is made in your browser and the file stays on your device.
Keep the file and the Proof of Time Certificate together.

A TimeBind shows the exact file existed no later than the recorded time and still matches. It does not show who made it, who owns it, or whether a watermark is present.

Make a TimeBind: first one free →

First Pooled TimeBind free, no card, then from US$5. An ASAP TimeBind is US$15.

What a TimeBind shows, and what it does not. It shows that an exact file existed no later than a recorded time and that a copy shown later still matches its fingerprint. It does not show who made the file, who owns it, that its contents are true, or whether AI was used to make or edit it. This is general information, not legal advice.

AI watermark and label decoder: FAQ

Can this tool detect a SynthID watermark in my file?

No. SynthID is a proprietary mark. Only Google's official SynthID Detector, or authorised partners, can check for it. This page explains what an official result, a C2PA manifest, a platform label, or a missing mark can and cannot show.

What is SynthID?

It is an invisible watermark developed by Google DeepMind and applied by some generators to images, video, audio and text at the time they are made. TIME named it one of the Best Inventions of 2026. Google has said it had watermarked more than 180 billion images and videos. A found mark is a provenance signal, not proof that the contents are true.

If the official detector finds no watermark, is the file real?

No. Many generated files never receive a SynthID mark, and a mark can be weakened or stripped. Google and the SynthID researchers have said a missing mark cannot establish human authorship.

How is this different from Content Credentials?

Content Credentials (C2PA) are a signed manifest some cameras, editors and generators attach to a file. SynthID is a hidden watermark some generators embed in the media itself. A file can have one, both, or neither. Our Content Credentials checker reads a C2PA manifest but does not verify the signature.

Do unofficial AI detectors work?

They are often wrong, including about real footage. Do not treat a third-party score as proof. Use official tools for the marks they were built to read, and keep a dated record of files you create.

What can a TimeBind add?

A TimeBind records the SHA-256 fingerprint of an exact file and the time. It shows that file existed no later than the recorded time and still matches. It does not detect watermarks, and it does not show who made the file or whether AI was used.

Related guides and tools

Your file’s fingerprint, anchored to a global network secured by real electricity and cryptography.

Editing a file—including with AI—changes its fingerprint, not its existing network record.

How this works

Your file’s SHA-256 fingerprint, a short code worked out from its exact contents, is made in your browser, and the file itself stays on your device. A cryptographic commitment to that fingerprint, which may be combined with others, is then anchored in a block on the Bitcoin blockchain. Bitcoin is secured by proof of work: computers around the world spend real electricity to add each block, and changing an earlier block would mean redoing that work faster than the rest of the network. Until that block is confirmed, your proof is pending. Once the Bitcoin anchoring is confirmed, you can verify the proof against that recorded commitment. The confirmed record can then be checked; Pooled TimeBinds are checked through TimeBinder, while ASAP TimeBinds can also be checked independently. If you check the file later and its fingerprint still matches, that shows the file existed no later than the recorded time. The record doesn’t show who made the file, whether AI was used, or that its contents are true.

Your filestays on your device
FingerprintSHA-256
Network timestampon Bitcoin

Date the original, with or without a watermark

Add a layer of protection against AI and human fraud and disputes: a dated, checkable record of the exact file you held.

Open the app and sign in with email or Google.
Add your files. They stay on your computer.
Seal them and download your Proof of Time certificates.

Open TimeBinder →

First Pooled TimeBind free, no card, then from US$5. See pricing.