AIHOT 于 2026-08-17 收录了“Anthropic 详解 Claude 隐形文本水印如何运作”这一公开动态。以下先呈现从来源页面抓取的正文，再给出 AIHOT 摘要与 TopoReduce 编辑解读。
PUBLIC SOURCE CONTENT
已抓取公开正文公开原文内容
Anthropic explains how Claude’s invisible text watermarks will work | The Verge
- AI
- News
- Tech
Anthropic explains how Claude’s invisible text watermarks will work
It’s using ‘a version’ of the open-source SynthID-Text system Google developed.
It’s using ‘a version’ of the open-source SynthID-Text system Google developed.
by Jess Weatherbed
Aug 17, 2026, 10:57 AM UTC
-
-
-
Image: Cath Virginia / The Verge, Getty Images
Jess Weatherbed is a news writer focused on creative industries, computing, and internet culture. Jess started her career at TechRadar, covering news and hardware reviews.
Anthropic has clarified how it’s planning to apply invisible watermarks to Claude-generated text in order to comply with Europe’s AI transparency rules. On Friday, Anthropic announced that Claude’s text marking system is “a version of the SynthID-Text approach” — an open-source watermarking technology developed by Google DeepMind that creates detectable patterns using wording probabilities.
This watermarking feature, alongside C2PA support for Claude-processed images, is being introduced to meet Anthropic’s obligations under the European Union’s AI Act, which requires synthetic audio, image, video, and text to include machine-readable marks that enable the content to be detected as artificially generated or manipulated. Anthropic says the text watermarks won’t make Claude more expensive for users, or “have any practical impact on the quality or content of Claude’s outputs.” Here’s Anthropic’s explanation for how it works:
Take the sentence “The weather today was cold and…”. The next word is very unlikely to be “sugary.” But it is quite likely to be “overcast” or “grey.” Under most circumstances, it doesn’t matter much to the reader which of these latter two words the model ultimately chooses—the meaning of the sentence is largely the same either way. In cases like this, the choice is settled by a random number.
Watermarking uses low-stakes choices like these—which occur many times over a piece of generated text—to leave a pattern in Claude’s responses. That pattern is undetectable to the reader, but is detectable to anyone who has a key that encodes it. When watermarking is used, choices are still made at random, but the source of the randomness is different. Instead of using an arbitrary random number generator to pick the next word, watermarking uses the key and a few words that come before to settle what word the model should pick.
As Anthropic notes, the EU’s AI transparency requirements also impact other major AI developers, so Claude won’t be the only model introducing text watermarks. Google’s Gemini chatbot has supported the SynthID-Text solution since 2024, and while OpenAI hasn’t detailed any text watermarking plans for ChatGPT in its AI Act compliance roadmap, it will also be subject to the law’s requirements.
Follow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates.
- Jess Weatherbed
- AI
- Anthropic
- News
- Tech
Most Popular
Most Popular
- Elon Musk broke the FAA — Palantir is picking up the pieces
- Apple’s camera-equipped AirPods appear in leaked video
- Comcast is turning millions of its routers into motion detectors
- Coyote vs. Acme is even funnier because Warner Bros. Discovery tried to kill it
- YouTube is changing how it counts views to give the numbers a boost
Advertiser Content From
This is the title for the native ad
More in AI
Robin Williams’ Instagram account brought back to fight ‘AI abuse’
OpenAI lays out new security changes after its AI hacked Hugging Face
Firefox’s Smart Window promises a better AI browser
Google’s Pet Memory forgot who my cats are
ChatGPT is getting a dedicated mode for teens
Elon Musk broke the FAA — Palantir is picking up the pieces
Robin Williams’ Instagram account brought back to fight ‘AI abuse’
Emma RothAug 18
OpenAI lays out new security changes after its AI hacked Hugging Face
Jay PetersAug 18
Firefox’s Smart Window promises a better AI browser
Stevie BonifieldAug 18
Google’s Pet Memory forgot who my cats are
Jennifer Pattison TuohyAug 18
ChatGPT is getting a dedicated mode for teens
Robert HartAug 18
Elon Musk broke the FAA — Palantir is picking up the pieces
Darryl CampbellAug 18
Advertiser Content From
This is the title for the native ad
Top Stories
Aug 18
Elon Musk broke the FAA — Palantir is picking up the pieces
Aug 18
Coyote vs. Acme is even funnier because Warner Bros. Discovery tried to kill it
Aug 18
ABC sues the FCC over Trump and Carr’s campaign of threats
Aug 18
Fairphone’s latest repairable phone is going on sale in the US
Aug 17
Trump’s border wall push is back — and it’s destroying national parks and 200-year-old trees
- News
- Tech
Anthropic explains how Claude’s invisible text watermarks will work
It’s using ‘a version’ of the open-source SynthID-Text system Google developed.
It’s using ‘a version’ of the open-source SynthID-Text system Google developed.
by Jess Weatherbed
Aug 17, 2026, 10:57 AM UTC
-
-
-
Image: Cath Virginia / The Verge, Getty Images
Jess Weatherbed is a news writer focused on creative industries, computing, and internet culture. Jess started her career at TechRadar, covering news and hardware reviews.
Anthropic has clarified how it’s planning to apply invisible watermarks to Claude-generated text in order to comply with Europe’s AI transparency rules. On Friday, Anthropic announced that Claude’s text marking system is “a version of the SynthID-Text approach” — an open-source watermarking technology developed by Google DeepMind that creates detectable patterns using wording probabilities.
This watermarking feature, alongside C2PA support for Claude-processed images, is being introduced to meet Anthropic’s obligations under the European Union’s AI Act, which requires synthetic audio, image, video, and text to include machine-readable marks that enable the content to be detected as artificially generated or manipulated. Anthropic says the text watermarks won’t make Claude more expensive for users, or “have any practical impact on the quality or content of Claude’s outputs.” Here’s Anthropic’s explanation for how it works:
Take the sentence “The weather today was cold and…”. The next word is very unlikely to be “sugary.” But it is quite likely to be “overcast” or “grey.” Under most circumstances, it doesn’t matter much to the reader which of these latter two words the model ultimately chooses—the meaning of the sentence is largely the same either way. In cases like this, the choice is settled by a random number.
Watermarking uses low-stakes choices like these—which occur many times over a piece of generated text—to leave a pattern in Claude’s responses. That pattern is undetectable to the reader, but is detectable to anyone who has a key that encodes it. When watermarking is used, choices are still made at random, but the source of the randomness is different. Instead of using an arbitrary random number generator to pick the next word, watermarking uses the key and a few words that come before to settle what word the model should pick.
As Anthropic notes, the EU’s AI transparency requirements also impact other major AI developers, so Claude won’t be the only model introducing text watermarks. Google’s Gemini chatbot has supported the SynthID-Text solution since 2024, and while OpenAI hasn’t detailed any text watermarking plans for ChatGPT in its AI Act compliance roadmap, it will also be subject to the law’s requirements.
Follow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates.
- Jess Weatherbed
- AI
- Anthropic
- News
- Tech
Most Popular
Most Popular
- Elon Musk broke the FAA — Palantir is picking up the pieces
- Apple’s camera-equipped AirPods appear in leaked video
- Comcast is turning millions of its routers into motion detectors
- Coyote vs. Acme is even funnier because Warner Bros. Discovery tried to kill it
- YouTube is changing how it counts views to give the numbers a boost
Advertiser Content From
This is the title for the native ad
More in AI
Robin Williams’ Instagram account brought back to fight ‘AI abuse’
OpenAI lays out new security changes after its AI hacked Hugging Face
Firefox’s Smart Window promises a better AI browser
Google’s Pet Memory forgot who my cats are
ChatGPT is getting a dedicated mode for teens
Elon Musk broke the FAA — Palantir is picking up the pieces
Robin Williams’ Instagram account brought back to fight ‘AI abuse’
Emma RothAug 18
OpenAI lays out new security changes after its AI hacked Hugging Face
Jay PetersAug 18
Firefox’s Smart Window promises a better AI browser
Stevie BonifieldAug 18
Google’s Pet Memory forgot who my cats are
Jennifer Pattison TuohyAug 18
ChatGPT is getting a dedicated mode for teens
Robert HartAug 18
Elon Musk broke the FAA — Palantir is picking up the pieces
Darryl CampbellAug 18
Advertiser Content From
This is the title for the native ad
Top Stories
Aug 18
Elon Musk broke the FAA — Palantir is picking up the pieces
Aug 18
Coyote vs. Acme is even funnier because Warner Bros. Discovery tried to kill it
Aug 18
ABC sues the FCC over Trump and Carr’s campaign of threats
Aug 18
Fairphone’s latest repairable phone is going on sale in the US
Aug 17
Trump’s border wall push is back — and it’s destroying national parks and 200-year-old trees
AIHOT 摘要
Anthropic 为遵守欧盟《人工智能法案》透明度要求，将采用 Google DeepMind 开源 SynthID-Text 技术的“一个版本”为 Claude 生成文本添加隐形水印。该水印利用低风险词汇选择的随机性模式标记内容，读者无感但持有密钥者可检测，且不会增加用户成本或影响输出质量。Anthropic 同时为 Claude 处理的图片引入 C2PA 支持。
为什么值得关注
水印机制把等义候选词的选择随机性替换为密钥编码，在不改变语义与成本的前提下提供机器可读标记，为评估欧盟透明合规对模型输出的实际影响提供具体参照。
工程化解读
从 TopoReduce 的工程视角看，这条信息属于“其他 AI 动态”主题。它的价值不只在于一个新产品或新观点本身，还在于说明 AI 系统正在如何影响模型接入、智能体协作、研发流程、基础设施和团队决策。实际采用前，应结合原文确认版本、适用范围、价格和运行条件。
- 发布时间：2026-08-17；AIHOT 分类：其他 AI 动态。
- AIHOT 标签：
- AIHOT 判断：水印机制把等义候选词的选择随机性替换为密钥编码，在不改变语义与成本的前提下提供机器可读标记，为评估欧盟透明合规对模型输出的实际影响提供具体参照。
- AIHOT 评分：50；评分用于站内排序，不等同于独立评测结论。
TopoReduce 编辑观察
当 AI 动态进入真实生产环境，团队需要同时关注能力边界、数据来源、调用成本、权限控制和可回滚性。把单条新闻放回完整工程链路中阅读，比只看标题更有助于判断它是否适合自己的产品和工作流。