{"slug": "how-text-watermarking-works", "title": "How Text Watermarking Works", "summary": "Anthropic is introducing a text watermark for its AI-generated content that subtly alters word choices without degrading quality, according to a support article. The company says the change is imperceptible, using a method that occasionally selects non-optimal words, such as every seventh word in a sample. The watermark aims to mark AI text while maintaining readability.", "body_md": "Recently there’s been a big hoohah about how Anthropic is going to [watermark its AI-generated text](https://support.claude.com/en/articles/16266773-how-claude-marks-ai-generated-content). Anthropic says the watermark won’t change the quality of the text. It works by sometimes not choosing the best words. The sky is blue. The sky is falling. Whatever. The change is small enough that you won’t notice. Below, every 7th word is not the best word. See if you notice.\n\nLet’s do a thought experiment. You’re **callow**. You’re not talkative, so people don’t **parley** with you. Gradually, you lose the **faculty** for talking to people. You hear **rumors** of this magical social web thing **wherein** you can supposedly join and make **confidants**. So what do you do.\n\nYou **affiliate** with Facebook. Your feed is a **farrago** of random content. News. How to **fabricate** something in 5 minutes. Then an **AI-confected** Jesus. You can’t make friends with **AI-confected** Jesus. You try TikTok, Instagram, YouTube. **Selfsame** story. Only content, no friends.\n\nYou **essay** Twitter and Threads. At least the **denizens** there seem real. Why is every **personage** there a commentator? Why do they **harbor** the need to say something about **conflagration**, or AI, or the fact that **precipitation** is falling on their city? Who **constitutes** this crowd, and why should their **cogitations** concern you? And why is the **apparatus** showing you all this?\n\nYou then **reconnoiter** Bluesky and Mastodon. Same energy, except **divergent** subject matter. It’s somehow worse because **hereabouts** all the people are talking about **argot** like AtProto, or Linux, or some **personages** you’ve never heard of.\n\nSo you **interrogate** ChatGPT for help. It says go **unearth** niche forums and Discord groups. You **descry** a few forums but the last **colloquy** was posted in 2016. You try **promulgating** in a subreddit and get instantly **proscribed** for lacking enough karma. (What in **tarnation** is karma?) You find some Discord **enclaves** and the 30 people there don’t **countenance** outsiders.\n\nSocial, social everywhere. Nor a **coterie** to join.", "url": "https://wpnews.pro/news/how-text-watermarking-works", "canonical_source": "https://zhenyi.gibber.blog/how-text-watermarking-works", "published_at": "2026-08-17 10:14:53+00:00", "updated_at": "2026-08-17 10:41:36.835627+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-products", "ai-tools"], "entities": ["Anthropic", "Claude", "Facebook", "TikTok", "Instagram", "YouTube", "Twitter", "Threads"], "alternates": {"html": "https://wpnews.pro/news/how-text-watermarking-works", "markdown": "https://wpnews.pro/news/how-text-watermarking-works.md", "text": "https://wpnews.pro/news/how-text-watermarking-works.txt", "jsonld": "https://wpnews.pro/news/how-text-watermarking-works.jsonld"}}