Claude users cancel subscriptions over Anthropic’s new AI watermark

Anthropic's decision to add an invisible watermark to text generated or processed by its Claude AI assistant has prompted some users to cancel their subscriptions, amid concerns about its impact on professional and academic work, Business Insider writes.

The AI company said some versions of Claude now embed an "imperceptible watermark" in text output.

The marker is designed to remain when content is copied and pasted and may survive some editing.

Anthropic says the watermark does not alter the meaning, quality or readability of responses. It has also stressed that a watermark indicates Claude processed text, rather than proving that the AI originally wrote it.

However, several users said the change made them uncomfortable about using Claude for work.

Arturo Villarroya, who works in public policy, said he cancelled his Claude Max subscription on Wednesday after becoming "uncomfortable" with the new system.

"I do use it for everything, mostly coding apps and tools," he said, adding that he also used Claude for sourcing, research and corrections.

Mr Villarroya said he was concerned that a watermark could remain when Claude was used only to translate text, correct spelling, fix academic citations or lightly summarise material.

Vladislav Rajtmajer, a freelance developer in the Czech Republic, said he cancelled his Claude Max subscription on Tuesday, citing the watermark as the main reason.

He said he used AI for code reviews and translations and was concerned that a marker on code delivered to clients could raise questions about authorship or lead to contractual penalties.

He added that the change was a "signal to look at Chinese AI providers and learn to work with other models."

Richard Echols, an AI consultant in Georgia, also said he cancelled his Claude Max subscription, with the watermark playing a major role in his decision.

"Even if you edit your own work that you wrote, they add the watermark anyway," Echols said.

Stevan Bogosavljević, a senior software engineer in Montenegro, said the watermark further reinforced his decision to cancel his subscription, after concerns about service quality, model access and stronger alternatives.

Having code attributed to Claude "definitely would be not desirable," he said, adding that an invisible marker could raise questions about authorship, compliance or client policies.

Anthropic said it has not seen an increase in cancellations since announcing the watermark.

The company says the technology is intended to improve transparency around AI-generated material and was introduced to comply with the European Union's AI Act. It plans to release a free API allowing users and third parties to check text for Claude's watermark.

Supporters argue that AI watermarks could make it easier to identify AI-generated material as its use becomes more widespread. Anthropic is not alone in using such technology: Google has its SynthID system, while OpenAI uses SynthID for supported images and audio. X also applies a "Made with AI" label to some content.

For some users, however, the issue is about control over their AI-dependent workflows.

Santiago Alessandro Rivera Martínez, a software engineer and founder of the digital agency Directa, said the watermark reinforced his concerns about becoming too dependent on Claude.

"When your workflow is built entirely around proprietary software, the provider can change the rules, features, restrictions, or conditions around that workflow later," he said.

"I would rather have the AI layer of my work be portable and provider-agnostic," he added.

The reaction comes as AI assistants become increasingly embedded in professional workflows, intensifying questions over authorship, disclosure and the use of AI.

By Aghakazim Guliyev