You can’t check this for an individual ChatGPT account. OpenAI’s provenance guidance says it is implementing invisible text watermarking for ChatGPT users in the EU. It doesn’t offer a consumer setting, status page or eligibility checker that confirms whether your account, or a given answer, carries a watermark. The only watermark switch OpenAI documents is for API customers. What you can do is work out which situation you are in, and how much a watermark result would tell you.
Which route applies to you
| Your situation | What OpenAI documents | What you can check |
|---|---|---|
| Regular ChatGPT account | Text watermarking is being implemented for ChatGPT users in the EU. Coverage can vary by product, model, export path, file type and when the content was created. | Read OpenAI’s live Help Center article, “Provenance signals in OpenAI-generated content.” There is no per-account lookup. |
| OpenAI API project or organization | Opt-in watermarking for supported models, available to API customers worldwide. | The Text provenance settings show which models are currently supported. |
| Research or academic organization | Text detector access, reviewed case by case. | Apply to OpenAI for access. Qualifying uses include provenance or detection-reliability work. |
ChatGPT consumer accounts
OpenAI’s article doesn’t tell individual users to open a particular ChatGPT setting. It also doesn’t give a “check my account” result. Don’t read that as a hidden toggle you’ve missed; no consumer toggle is documented.
Being in the EU doesn’t mean every output is watermarked either. The guidance says coverage varies by product, model, export path, file type and creation date. If the date or region of your use matters, treat the rollout as something to confirm against the live article, because these details change.
OpenAI API: turn watermarking on
Watermarking is opt-in for API customers. These steps apply only to the API platform, not to ChatGPT.
Recommended Free Tools
#1 Best Overall
- ✅ PROTECT ONLINE ACCOUNTS – A password manager, two-factor security key, and secure communication token in one, OnlyKey can keep your accounts safe even if your computer or a website is compromised. OnlyKey is open source, verified, and trustworthy.
- ✅ UNIVERSALLY SUPPORTED – Works with all websites including Twitter, Facebook, GitHub, and Google. Onlykey supports multiple methods of two-factor authentication including FIDO2 / U2F, Yubico OTP, TOTP, Challenge-response.
- ✅ PORTABLE PROTECTION – Extremely durable, waterproof, and tamper resistant design allows you to take your OnlyKey with you everywhere.
- ✅ PIN PROTECTED – The PIN used to unlock OnlyKey is entered directly on it. This means that if this device is stolen, data remains secure, after 10 failed attempts to unlock all data is securely erased.
- ✅ EASY LOG IN –No need to remember multiple passwords because by plugging OnlyKey to your computer, it automatically inputs your username and password. It works with Windows, Mac OS, Linux, or Chromebook, just press a button to login securely!
For one project
- Open the project’s Settings.
- Go to Text provenance.
- Turn on Allow text watermarking.
- Select the available model or models, then save.
For organization-wide defaults
- Open Organization settings.
- Go to Data controls → Text provenance.
- Turn on Allow text watermarking, select models and save.
The model list shown on that screen is the current one, and OpenAI says it is extending coverage over time. If a model you use isn’t listed, it can’t be watermarked yet.
Detector access is separate
Enabling watermarking in the API doesn’t give you the detector. OpenAI limits its text detector to approved research and academic organizations and other qualifying uses, with applications reviewed case by case. An individual who wants to paste in a ChatGPT answer and see a result has no general route to do so.
Rank #2
- POWERFUL SECURITY KEY: The Security Key C NFC is the essential physical passkey for protecting your digital life from phishing attacks. It ensures only you can access your accounts.
- WORKS WITH 1000+ ACCOUNTS: Compatible with Google, Microsoft, and Apple. A single Security Key C NFC secures 100 of your favorite accounts, including email, password managers, and more.
- FAST & CONVENIENT LOGIN: Plug in your Security Key C NFC via USB-C and tap it, or tap it against your phone (NFC) to authenticate. No batteries, no internet connection, and no extra fees required.
- TRUSTED PASSKEY TECHNOLOGY: Uses the latest passkey standards (FIDO2/WebAuthn & FIDO U2F) but does not support One-Time Passwords. For complex needs, check out the YubiKey 5 Series.
- BUILT TO LAST: Made from tough, waterproof, and crush-resistant materials. Manufactured in Sweden and programmed in the USA with the highest security standards.
What the watermark is
OpenAI calls its technique textGrain. It subtly shifts the statistical pattern of the model’s choices of words and word pieces. It doesn’t insert hidden characters, invisible spaces or odd punctuation, so there is nothing to find by inspecting the text or its characters.
A detector tests whether a passage follows a secret pattern more often than chance would predict. Its output is a provenance signal. OpenAI says a signal can suggest that an OpenAI model generated or processed the text. It cannot establish:
Rank #3
- Requires 3 "AAA" batteries (included)
- Unit auto-locks for 30 minutes after 5 consecutive incorrect PINs
- who used the model;
- whether a human contributed, or how much text the model produced;
- who owns the content;
- whether the content is accurate.
In OpenAI’s words: “Provenance signals can provide useful information about origin, but they do not guarantee that content is accurate, unedited, legally owned, or presented in the correct context.”
When detection is unreliable
- Short text: there may not be enough evidence. OpenAI says the EU AI Act Code of Practice doesn’t require watermarking in outputs under 200 tokens (about 150 words in English) or in code snippets. That is a floor, not a promise that longer prose is always watermarked or detectable.
- Code: fewer plausible next-token choices leave little room for a signal.
- Factual or exact-reproduction text: constrained wording limits the signal in the same way.
- Language: detection varies by language.
- Editing: rewriting, paraphrasing or translation can reduce reliability. Keeping the original wording may preserve the signal but doesn’t guarantee detection.
OpenAI’s language chart (2026) reports detection rates at a 1% false-positive rate. It used 500 synthetic English prompts translated into the other 23 official EU languages. Spanish scored highest at 69.0% and Romanian lowest at 42.2%. OpenAI says it raised watermark strength for languages below 60%, shown in the chart’s light-blue portions. These figures come from that evaluation setup and aren’t a guarantee for text you submit.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Reading a result, or the lack of one
A missing signal doesn’t prove a human wrote the text or that it didn’t come from OpenAI. OpenAI lists unsupported products, models or exports, absent signals and degraded signals as possible reasons. Third-party AI classifiers don’t help here either: OpenAI distinguishes their pattern-based guesses from its watermark detector, so their scores don’t verify an embedded watermark.
Don’t ask ChatGPT whether it wrote something. OpenAI’s FAQ says: “ChatGPT has no ‘knowledge’ of what content could be AI-generated or what it generated.”
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallCrashes, 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 minuteBest Value
- FIDO-ONLY FUNCTIONALITY: Supports FIDO2 (passkeys) and FIDO U2F protocols for passwordless and second-factor authentication. Does not support OTP, TOTP, Smart Card (PIV), or other advanced features - upgrade to YubiKey 5 Series for extended functionality
- SECURE AND CONVENIENT: Passwordless MFA login with the YubiKey Bio authenticator and biometric information using a fingerprint, with a PIN as a fallback. Simply plug in via USB and use your fingerprint to authenticate
- DEVICE & OS COMPATIBILITY: Compatible with Windows, macOS, ChromeOS, and Linux. Works seamlessly with supported services like Google and Microsoft accounts, and major password managers. See the full compatibility list at "Works With YubiKey"
- DURABLE & RELIABLE: Resistant to tampering, water, and crushing. No batteries or network connectivity required, offering dependable authentication without any downtime. Securely manufactured in USA & Sweden
- Yubico Authenticator App - Fingerprint enrollment, passkey management and PIN configuration available via the app app - Upgrade to YubiKey 5 Series to generate one-time-passwords (OTP) via Yubico Authenticator and for advanced compatibility (OATH, PIV)
Practical checklist
- Decide whether you use ChatGPT or the API. Only the API has a documented watermark setting.
- For ChatGPT, read OpenAI’s live provenance article for current regional scope. Don’t assume your account, model or export is covered.
- For the API, open the Text provenance settings and confirm your model is listed before relying on watermarking.
- If you need detection, check whether your organization qualifies to apply for detector access.
- Treat any outcome as a provenance signal, not an authorship verdict.
OpenAI’s rollout and supported-model details were last confirmed on 7 October 2026 and can change. The article’s broader EU context sits in OpenAI’s “EU AI Act: OpenAI Resources and Customer Guidance” page.
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.
