ChatGPTの文章に見えない透かし、EUで導入し判定の限界も公表

"October 5, 2026"（公開日）／"Over the coming weeks, we will introduce text watermarking to eligible ChatGPT and Codex users across all plans in the EU only. We are not making text watermarking a global default at launch. This regional approach gives us room to learn from real-world use and feedback."（Our next steps）— OpenAI 公式発表より

"Starting today, API customers globally will be able to opt in to text watermarking for select models. Text watermarking will remain off by default in the API."／"We are also working with cloud partners to make watermarking available for OpenAI model outputs accessed through their services in the coming weeks."（Our next steps）— OpenAI 公式発表より

"Approved researchers and expert organizations can apply starting today. In accordance with the Code of Practice, access will be initially granted on a case-by-case basis to support evaluation and improvement of text provenance. The tool will report whether it detects an OpenAI watermark, without identifying the user or revealing their prompts or conversations. Given the risk of missed watermarks and false positives, we are not making it publicly available at launch."（Our next steps）— OpenAI 公式発表より

"A language model generates text one token at a time, sampling each token (a word or part of a word) from a next token prediction (NTP) distribution determined by the preceding text. Watermarking embeds a statistical signal by modifying this sampling using pseudorandom values derived from a secret key and the preceding context"（1 Introduction）／"The detector requires only the generated text and the secret key"（Abstract）— OpenAI 技術レポートより

"Across the benchmarks we use to assess Astra, our latest frontier model, we do not see meaningful performance differences with and without watermarking."（Impact of watermarking on output quality）— OpenAI 公式発表より

"In our evaluations, textGrain matched or exceeded the performance of other approaches we tested, including SynthID for text. Even so, strong performance under ideal conditions does not guarantee reliable detection in everyday use."（How our watermarking works and performs）— OpenAI 公式発表より

"Shorter or more constrained text is harder to detect. At a target false positive rate of 1%, our detector identified watermarks in about 80% of 200-token passages, compared with about 95% of 400-token passages, for content such as psychology. Detection rates were substantially lower for content such as mathematics, where there is less flexibility in word choice."（How our watermarking works and performs）— OpenAI 公式発表より

"Editing can weaken the watermark. In an evaluation of 400-token passages, replacing 10% of words with synonyms reduced detection from about 92% to 66%. Replacing 25% of words reduced it to 17%."／"Results are based on watermarked English responses to questions from ELI5."（How our watermarking works and performs）— OpenAI 公式発表より

"A watermark does not measure human contribution. It can indicate that an OpenAI system generated or processed part of a passage, but not how much human judgment, editing, or creativity went into it."／"The absence of a detected watermark does not prove human authorship. Text generated with OpenAI tools may be too short, edited, or translated for detection to work reliably. It may also come from an unsupported model, predate watermarking, or have been generated by another company's tools."（What a text watermark doesn't tell you）— OpenAI 公式発表より

"Providers of AI systems, including general-purpose AI systems, generating synthetic audio, image, video or text content, shall ensure that the outputs of the AI system ／ are marked in a machine-readable format and detectable as artificially generated or manipulated."（Article 50(2)）— EU AI法 条文より

"Transparency rules (Article 50) start to apply."（02 Aug 2026）／"Transitional deadline for certain providers of AI systems, including GPAI systems, generating synthetic content, that were already placed on the market before 02 Aug 2026, to comply with Article 50(2)."（02 Dec 2026）— 欧州委員会 AI Act Service Desk より

"Our teams have developed a text watermarking method that we continue to consider as we research alternatives."／"it is less robust against globalized tampering; like using translation systems, rewording with another generative model, or asking the model to insert a special character in between every word and then deleting that character - making it trivial to circumvention by bad actors."／"For example, it could stigmatize use of AI as a useful writing tool for non-native English speakers."（Text watermarking）— OpenAI 公式発表より