Anthropic faces backlash as Claude users cancel subscriptions over new AI watermark

Published Aug 14, 06:48 PM UTC

Anthropic's decision to embed an imperceptible watermark in text generated by its Claude AI assistant has sparked backlash, with dozens of users on X and at least four interviewed by Business Insider canceling their subscriptions. The watermark, applied globally since August 2, is designed to distinguish AI-generated content from human writing and may survive editing. Users, including developers and policy professionals, fear the mark could appear in client work such as code or documentation, raising authorship questions or triggering contract penalties. Some are switching to alternatives like Cursor, Grok, or Chinese AI providers. Anthropic stated it has not seen a trend of increased cancellations and emphasized that the watermark indicates Claude processed the text, not necessarily that it authored it. The company has 300,000 business customers as of September 2025. Supporters argue watermarks improve transparency and help prevent AI model degradation.