![\"A](\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/08\/d41586-026-02235-8_53042962.jpg\"\/)
<\/p>\n <\/p>\n
An AI fact-checking tool found errors in molecule boiling points listed in a chemistry reference database.Credit: Monty Rakusen\/Getty<\/p>\n
For decades, chemists have relied on handbook values for a molecule\u2019s boiling point to identify substances and plan processes such as distillation. But an artificial-intelligence model has revealed that some trusted numbers in one reference database have been wrong all along.<\/p>\n
Sebastian Pios, a theoretical chemist at Zhejiang Lab in Hangzhou, China, was using an AI system to predict the boiling points of several molecules when it began producing values that clashed with long-accepted entries in a 75-year-old reference database. At first, he thought the model was wrong. But when he manually checked the original literature, he found that the reference data were wrong, not the AI model.<\/p>\n
In two other cases, Pios\u2019s AI model spotted errors in older papers and reference books \u2014 mistakes that have made their way into the scientific canon. One was a typo in a paper; another was incorrect values of century-old boiling-point measurements. Both errors are likely to have caused researchers using the database \u201ca lot of trouble\u201d, says Pios.<\/p>\n
Pios is one of a growing number of scientists using AI as a tool for auditing scientific knowledge. As well as checking databases, researchers are using specialized AI tools to hunt for errors in papers published in journals and conferences<\/a>.<\/p>\n AI, peer review and the human activity of science<\/p>\n <\/a><\/p>\n In an analysis posted online on 22 July<\/a>, researchers at SAI Labs, a for-profit research-review company in Delaware, used AI agents to assess 168 papers selected for oral presentation at the 2026 International Conference on Machine Learning (ICML). The AI agents extracted the authors\u2019 central claims, downloaded accompanying resources, reran experiments where possible and compared the results with those reported by the authors.<\/p>\n Of the 92 papers that had at least five claims available for assessment, the AI agents were able to reproduce more than two of the five claims from only 34 papers. The agents successfully repeated more than 80% of claims from just eight papers.<\/p>\n But such AI fact-checking tools remain unreliable arbiters of the scientific corpus, says Odd Erik Gundersen, a computer scientist at the Norwegian University of Science and Technology in Trondheim. The tools \u201cmake mistakes like humans do\u201d, he says, which is why the quality of AI fact checkers must be manually processed with human oversight.<\/p>\n AI fact checker<\/p>\n Many researchers already dedicate their time to spotting errors in papers and use tools to check certain facets of papers<\/a>. But one advantage of using AI is the speed at which it can scan scientific databases and literature compared to humans, says James Zou, a computer scientist at Stanford University, California. \u201cThe biggest difference is to be able to do this at a scale that was not possible before.\u201d<\/p>\n In a study posted on the preprint server arXiv1<\/a>, Zou and his colleagues used an \u2018AI checker\u2019 to scan papers published at NeurIPS \u2014 a prestigious annual AI research conference \u2014 for errors. Their tool found that errors in papers rose from 3.8 in 2021 to 5.9 in 2025 \u2014 an increase of 55%.<\/p>\n![\"\"](\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/08\/d41586-026-02235-8_51208868.png\"\/)
<\/p>\n
![\"\"](\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/08\/d41586-026-02235-8_50837996.png\"\/)
<\/p>\n