Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →OpenAI says it will begin adding an invisible statistical watermark, called textGrain, to eligible ChatGPT and Codex text outputs in the European Union over the coming weeks. The announced rollout covers eligible users on all plans, but it is not yet described as complete. Outside that rollout, API customers worldwide can opt in for select models; watermarking is off by default for API use.
What OpenAI is changing—and where
OpenAI announced the plan on October 5, 2026, saying it will add textGrain to eligible ChatGPT and Codex text outputs in the EU. The company says the rollout will happen over the coming weeks, so the announcement does not mean every eligible account already receives watermarked output. OpenAI is not making the feature a global default at launch. OpenAI’s announcement describes the rollout.
As an Amazon Associate I earn from qualifying purchases.
| Channel | Availability and default |
|---|---|
| ChatGPT and Codex | Eligible text outputs in the EU, across all plans; rollout announced for the weeks after October 5, 2026. OpenAI does not say it is complete. OpenAI |
| API | Customers globally can opt in for select models. Watermarking is off by default, and the supported model list may change. OpenAI |
| Text detector | Initially limited to approved researchers and expert organizations, with applications reviewed case by case. OpenAI |
OpenAI’s public image and audio verification tools are separate; the company’s announcement does not say that the public tools detect text watermarks. OpenAI Help Center
How textGrain works
OpenAI describes textGrain as a statistical pattern in a model’s word or word-piece choices. It subtly influences those choices so a detector can look for the resulting pattern. It does not insert a special symbol, invisible space, unusual punctuation, or watermark-only token, so copying and pasting text will not expose a hidden character string. OpenAI Help Center
#1 Best Overall
What the EU rules require—and what they do not
The EU policy context is broader than OpenAI’s particular implementation. The European Commission says the AI Act’s transparency obligations began applying on August 2, 2026. They include requirements for providers to design systems to inform people when they interact directly with AI and to add machine-readable marks that help detect AI-generated or manipulated content. The Commission also describes disclosure duties for deployers in specified cases, including deepfakes and certain AI-generated or manipulated text on matters of public interest. European Commission, July 20, 2026
The Commission’s Code of Practice is voluntary and provides practical steps for meeting the Act’s transparency obligations. The Commission says clear labelling is required in key cases, including deepfakes and AI-generated or manipulated text published on matters of public interest. European Commission, June 10, 2026 OpenAI says its rollout responds to the Act and its commitments under the Code. That does not mean the law requires textGrain specifically or that every AI text system must use the same watermarking method.
Rank #2
Short passages, code, and constrained writing are harder cases
OpenAI’s Help Center says the Code of Practice does not require watermarks in outputs shorter than 200 tokens—about 150 English words—or in code snippets. That is distinct from whether a passage can be detected: short text may not contain enough signal, while code and other constrained subjects offer fewer plausible word choices for the watermark to influence. OpenAI Help Center
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →OpenAI’s own evaluation reported lower detection for mathematics than for psychology passages, attributing the difference to less flexibility in word choice. The figures below are company evaluation results, not independent real-world guarantees; they should not be generalized to all languages, models, passage types, or editing patterns. OpenAI
Rank #3
| OpenAI-reported test | Reported result |
|---|---|
| Psychology passages of 200 tokens, at a target 1% false-positive rate | About 80% detection. OpenAI evaluation, 2026. OpenAI |
| Psychology passages of 400 tokens, at a target 1% false-positive rate | About 95% detection. OpenAI evaluation, 2026. OpenAI |
| Baseline detection, compared with replacing 10% of words with synonyms | About 92% at baseline, falling to 66% after the synonym replacements. OpenAI evaluation, 2026. OpenAI |
| Replacing 25% of words with synonyms | 17% detection. OpenAI evaluation, 2026. OpenAI |
OpenAI cautions that performance under ideal test conditions does not guarantee reliable detection in everyday use.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What a watermark result can—and cannot—show
A positive detector result is evidence of a detected OpenAI watermark, not a complete account of how a passage was written. OpenAI says a watermark does not measure human contribution: it cannot show how much a person edited, judged, or shaped the text. Nor does it identify a user, reveal prompts or conversations, determine ownership or legal responsibility, or establish whether the passage is true or misleading. It is not a plagiarism detector or a definitive authorship test. OpenAI Help Center
A negative result is not proof that a person wrote the text. OpenAI says short, edited, or translated passages may evade detection, as may output from unsupported or older models and systems from other providers. The detector is initially restricted to approved researchers and expert organizations; OpenAI says its result indicates whether an OpenAI watermark was detected without revealing a user, prompts, or conversations. OpenAI
Quick Recap
Best Value
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.
