Josh Goolnik·2d agoAI Analogies 10: Text WatermarkingYou might have heard recently that Anthropic (makers of Claude, which is a popular Large Language Model or LLM) has started to watermark…
Art Morales, PhD·Oct 25Hidden Watermarks in AI-Written ScienceLike everyone keeping with the times, I use AI to help draft and organize my ideas into more formal writing, and since August every new…
InThe Post-Project WorldbyLuigi Pascal Rondanini·Oct 24AI Watermark Watch — Week 9: What If the Watermark Can Be Defeated Without Knowing Its Secret?New independent research identifies a shared weakness across seven statistical watermarking methods, including SynthID. The implications…
InThe Post-Project WorldbyLuigi Pascal Rondanini·Oct 11AI Watermark Watch — Week 8: What If the Watermark Makes the AI Less Accurate?New research raises an uncomfortable possibility: a technology designed to establish AI provenance may also introduce factual errors into…
Akurindo W·Oct 8AI Watermarking Has Arrived — Don’t Confess, Document ItPicture the email landing in your inbox: “Hi, quick question — was this draft written by AI?” Attached is a screenshot of a detection…
Eleventh Hour Enthusiast·Sep 27The Signal in the SamplerHow Claude’s Text Watermark Turns Token Choices into Evidence
InThe Post-Project WorldbyLuigi Pascal Rondanini·Sep 27AI Watermark Watch — Week 6: What If the Watermark Changes the AI?This week’s report has been researched and drafted by the AI agents assigned to AI Watermark Watch and verified by the human behind these…
Yogesh Yadav·Sep 19How to Build a Scalable Parallel Audio Watermarking System on AWS LambdaHow we split, watermark and re-join long audio in parallel with S3 events, Lambda and DynamoDB, and why knowing when every worker is done…
