Paris‑based tech entrepreneur Guillaume Meyer unexpectedly went viral after releasing his open‑source project “Watermarks Remover,” built in response to Anthropic’s invisible watermark announcement.
Curious about how watermarking worked, Meyer coded a tool in just five hours to disrupt statistical patterns used to flag AI‑generated text. Within days, his posts on X and LinkedIn drew millions of impressions, turning a side experiment into a full‑time commitment.
Watermark Remover Tool
Meyer argues that water marking creates more problems than it solves. He points out that tools like Grammarly, which rely on AI, could cause all user‑generated content to be labeled as AI‑authored, even if only minor edits are made.
He supports content attribution but believes current water marking techniques risk false positives and undermine trust.
The project, which rewrites text or pixels until watermarks are disrupted, has attracted global contributors. While Meyer welcomes the attention, he stresses the tool is imperfect and will require ongoing adjustments as detection methods evolve. He insists his intent is educational, not to enable impersonation or theft.
Now, Meyer is considering how to transform the viral project into a business, though legal and technical challenges remain. For him, the debate reflects broader unease with AI watermarking and the need for more thoughtful solutions.
