I created a viral AI watermark remover. I wasn't ready for all the attention.

Paris-based tech entrepreneur Guillaume Meyer created an open-source AI watermark remover tool after Anthropic announced its own AI watermark feature. Meyer's project, Watermarks Remover, gained viral attention on social media and GitHub after it was published. He believes AI watermarks create more problems than they solve, but supports content attribution.

Meyer's tool works by checking for watermarks at the beginning, generating slight variations of the text or pixels while preserving meaning, and checking again until the watermark is disrupted and no longer flagged. The tool can be applied to both text and images. Meyer built the project in just a few hours and has been working on it full-time since then.

The project quickly gained traction, with over 2 million impressions after Meyer's second X post about it. Meyer, who is careful about his privacy, was not prepared for the sudden attention and had to create new social media accounts to manage the influx of comments and messages. Most of the feedback was positive, but there were some negative comments from people who wanted to know when something is AI-generated and considered Meyer a bad person for not wanting to detect AI content.

Meyer is thinking about turning the project into a business, but is unsure about the legal feasibility and the technical difficulty of creating a simpler solution. He believes the conversation about AI watermarking is important and should have happened before the regulation came into effect, but is happy that it is happening now.