AI Detection in Enterprise Publishing and Editorial Workflows
Editors need clear rules for using AI detection as one signal in decisions, not as final judgment.
Soren Halvgaard
Section
AI Detection & Fingerprinting
7 stories in AI Detection & Fingerprinting.
AI Watermarking Standards and Their Limitations
Current watermarking standards cannot reliably detect AI-generated text after editing.
Miroslava Čech
Model Fingerprinting in Large Language Model Output
Detecting which model wrote text reveals why single-model pipelines need independent quality checks.
Priya Nandakumar
AI Content Detectors Rated for False Positive Risk on Human-Written Text
Detector scores alone cannot determine whether a human or AI wrote a given text.
Miroslava Čech
Why Multi-Model Output Evades Single-Model Detectors
Mixing multiple AI models breaks detectors trained on single-model fingerprints.
Soren Halvgaard
Perplexity and Burstiness as Detection Signals
Perplexity and burstiness measure different writing patterns that detectors use to flag AI output.
Miroslava Čech
How Accurate Are AI Detectors
Vendor claims of 95% accuracy don't match real-world performance in independent testing.
Priya Nandakumar