●researcherProvides a more realistic evaluation surface for AI-text detectors than binary human-vs-AI benchmarks.
●policyWorth watching because it quantifies how AI authorship signals accumulate through co-editing workflows, directly relevant to academic integrity and content provenance policy.