Topic
Post-publication scrutiny of published papers by independent sleuths, and publisher and journal handling of the resulting concerns.
An international team flagged 4,085 of 33,159 articles from 2012 to 2023 as potentially linked to paper mills, according to a PNAS study. The flagged papers mostly cite and are cited by each other, and that loop inflates citation metrics at some journals.
Reality
Mathematician Thomas Bloom stopped tagging Erdős problems as open or solved before OpenAI posted 722 AI-produced manuscripts. A DeepMind paper in Science shows the documented, human-vetted alternative: 9 of 353 open problems solved.
Reality
Nikon is re-reviewing its Small World in Motion winner after scientists said the microscope video shows structures that do not occur in biology. Its ruling on whether neural-network coloring counts as banned generative AI rests on documents the entrant supplied after he won.
Perspective Coverage
4 publishersReality
Nature found more than 30 academic press releases celebrating honours from sham societies that used scientists' names without consent. Several academies the societies named as partners told Nature they knew of no such ties, a check universities could have made before announcing anything.
Reality
A run of vulnerability, hacking and mathematics announcements from Anthropic and OpenAI drew fast coverage. The expert re-readings that followed described plagiarism accusations and basic security failures.
Reality
Elisabeth Bik's 2016 survey of more than 20,000 papers documented 782 image duplications. The publisher of Physics World now assigns six research-integrity officers to misconduct cases that were once rare in physics.
Reality
A Caltech open letter says research mathematicians end up verifying and correcting what AI companies announce, without pay or credit. OpenAI's sponsorship was $1m in credits, and it took them back.
Reality
The Fields Medalists behind a new open letter accept that AI can speed up mathematics. Their objection is to solutions announced before anyone has had time to write them up properly or work out whose ideas they used.
Reality
OpenAI says the run took 88 hours and cost millions, which tells a small research group more about whether it can compete than the proof itself does, and its wording on user data shows the limit of what any vendor can promise.
Reality
A UK molecular biologist says dozens of cell-ageing studies used an antibody against a bacterial protein to detect a mammalian one. It is his third reagent flag in about three months.
Reality