Anthropic’s Claude will watermark AI-generated text. Here’s how it works

Anthropic is implementing a new method to watermark AI-generated text, embedding a hidden digital signature within the content produced by its Claude models. This move is a direct response to the European Union's AI Act, which mandates transparency rules that took effect at the beginning of the month.

The technology works by adding a subtle, undetectable signal to the AI's output. This signal can later be detected to confirm the text originated from Anthropic's systems. The goal is to provide a reliable way to distinguish between human-written and AI-generated content, a crucial step in addressing the growing concerns surrounding the proliferation of artificial intelligence in various forms of communication and media.

By adhering to the EU's AI Act, Anthropic demonstrates its commitment to responsible AI development and deployment. The transparency requirements are designed to foster trust and accountability in the rapidly evolving field of artificial intelligence, ensuring users can be aware of the origin of the information they consume.

“

The move is part of Anthropic's commitments under the European Union's AI Act, which includes transparency rules that kicked in early this month.