Anthropic's Watermark Plan Leads Some Users to Cancel Claude

Anthropic's plan to add watermarks to Claude-generated text has led some users to cancel their subscriptions. According to Business Insider, four users who canceled are worried about permanent marks on their work that could reveal AI use in client projects. Anthropic says it hasn't seen a big drop in subscriptions.

Arturo Villaroya, who works in public policy, canceled his $100-per-month Claude Max plan. He uses Claude for almost everything, including coding and research. He's concerned that even using Claude for translation or proofreading could leave a mark, causing trouble in strict academic or professional settings.

Anthropic explains that a watermark only shows Claude processed the text, not that Claude wrote it, but this hasn't eased all worries. Developers like Vladislav Reitmair from Czechia and AI consultant Richard Echols from Georgia also canceled, fearing hidden marks on delivered code could lead to compliance issues.

Not everyone is against watermarks. Some believe they increase transparency and prevent AI systems from training on AI-generated content, which can degrade performance. Anthropic says it's adding watermarks to comply with the EU's AI Act and plans to offer a free API for third-party detection.