Skip to main content
OnTheRiceSEE THE WORLD EARLY.
Search the world…
OnTheRice’s Emerging DataTechnology

OpenAI’s text watermark has separate marking and checking gates

By OnTheRice · Sources and limitations · Report a correction

OpenAI separates the ability to add a text watermark from permission to check it. Teams adopting the feature still need their own record of AI use and human review.

Current priceNot available
24-hour changeNot available

Source confidenceNot assessed

Published
Check timeNot applicable

What’s changing?

Assign responsibility for each step rather than treating one product setting as a complete publication check.

Key points

  • OpenAI announced optional watermarking for selected API models on 5 October, off by default. Eligible EU ChatGPT and Codex output is due to follow over coming weeks.
  • Approved researchers and expert organisations can apply for detector access. Turning marking on does not grant that access.

What could change this view?

  • A watermark can be useful evidence of model involvement, but it does not identify the writer, measure human contribution or verify truth. Missing detection does not prove human authorship. No outputs or detector were independently tested here.

Bottom line

Publishers, editorial teams and developers using OpenAI text output. Action: Inspect the supported-model list and project or organisation setting. Keep records of AI assistance and human review, and assess disclosure requirements for the intended use. Only test text through authorised detector access.

This is based on the information recorded with this signal.

How likely is this call? · Three frameworks

This signal has no investible bullish or bearish call. A market-direction percentage does not apply.

Read the shared method · Read the public probability method

Evidence behind this view

Checks whether separate pieces of evidence support the same view.

Recorded connection

Assign responsibility for each step rather than treating one product setting as a complete publication check.

We have not established a separate count of independent evidence clusters for this card. Several reports may rely on the same event or data provider.

What past price moves show

Compares today’s price moves with recorded past patterns.

This card has no recorded investible call. A price-pattern comparison is not applicable unless a specific asset is added with evidence.

Evidence against this view

Shows evidence and risks that could challenge this view.

Recorded opposing evidence and risks

A watermark can be useful evidence of model involvement, but it does not identify the writer, measure human contribution or verify truth. Missing detection does not prove human authorship. No outputs or detector were independently tested here.

Why the direction changed

Explains recorded changes from expecting a rise to expecting a fall, or the reverse.

No documented bullish-to-bearish or bearish-to-bullish change is available for this card. We keep the original call and its time unchanged.

What people say vs the facts

Compares recorded public discussion with checked market facts.

No separate Reddit or public-discussion comparison with verified market data is recorded for this card. We cannot treat social attention as proof of a market move.

How past forecasts performed

Shows recorded forecast results and what they can teach us.

This card has no recorded investible forecast to score.

No tested mapping from evidence combinations to future success is recorded for this card. Confidence in sources and past outcomes do not establish a calibrated probability.

RETAINEDHistorical record
Sources, original wording and price checks

Published

First published through Sites 06 Oct 2026, 12:18:23 SGT

No explicit investible BULLISH or BEARISH call on this card.

What changed

Recorded public history

OpenAI separates the ability to add a text watermark from permission to check it. Teams adopting the feature still need their own record…

This is the first recorded public view. Further public updates will appear here.

  1. OpenAI’s text watermark has separate marking and checking gates

    OpenAI separates the ability to add a text watermark from permission to check it. Teams adopting the feature still need their own record of AI use and human review.

    First published by OTR: 6 Oct, 12:18 SGTPublished by OTR: 6 Oct 2026, 12:18 SGT
    Our approach to EU text provenance rules ↗Provenance signals in OpenAI-generated content ↗Guidelines on transparency obligations for providers and deployers of certain AI systems ↗

Why the view changed

This is the first recorded view. Future source-backed updates will appear here.

Evidence, memory & review 3 linked sources

Named assets or actors: OpenAI, privately held; no listed instrument assigned

Directional takeaway: Direction not assigned No source-grounded listed-asset return mechanism is established. NO ENTRY

OTR reasoning: Assign responsibility for each step rather than treating one product setting as a complete publication check.

Counterpoint: A watermark can be useful evidence of model involvement, but it does not identify the writer, measure human contribution or verify truth. Missing detection does not prove human authorship. No outputs or detector were independently tested here.

Source-grounded facts (paraphrased)

  • Time not supplied: OpenAI announced optional watermarking for selected API models on 5 October, off by default. Eligible EU ChatGPT and Codex output is due to follow over coming weeks. Source records: source-1, source-2, source-3
  • Time not supplied: Approved researchers and expert organisations can apply for detector access. Turning marking on does not grant that access. Source records: source-1, source-2, source-3

Evidence basis: Source-checked editorial summary; this is a paraphrase and synthesis, not a source quotation.

Source trail

Review status: CHIEF APPROVED AFTER SOURCE REVIEW

No score or probability is presented.