Anthropic to Launch Watermark Detection API for Claude AI Texts

Anthropic will soon launch a watermark detection API for its Claude AI text generation model. This technology builds on Google’s SynthID method, modifying randomness during word selection to maintain text quality. However, the approach has limitations, particularly with fact-heavy text, code, and extensive rewriting. This development follows Anthropic’s recent focus on enhancing Claude’s capabilities, including setting Claude’s code to auto mode by default for pro plans. The new API will allow third parties to verify the authenticity of texts generated by Claude, addressing growing concerns over AI-generated content integrity. For practitioners, this means enhanced tools for ensuring content authenticity in applications relying on Claude’s outputs. The Decoder reported.