Scientific AI systems require domain-aware robustness validation

AI has become an essential tool in many areas of scientific discovery. However, its robustness, particularly in high-stakes scientific applications, remains insufficiently understood. While adversarial vulnerabilities have been widely studied in image recognition, their implications in scientific disciplines, where physical constraints and numerical stability are critical, have not been thoroughly explored. This work presents a systematic evaluation of AI model fragility across five scientific domains: climate modeling, quantum chemistry, fluid dynamics, quantum chromodynamics, and wireless communications. We demonstrate that small, structured perturbations can destabilize predictions in all of these domains, even in models that achieve high accuracy under normal conditions. Furthermore, the failure patterns vary across domains, suggesting that these vulnerabilities are domain specific. Our findings underscore the importance of assessing AI robustness in scientific contexts and offer preliminary insights into how to design more resilient and reliable AI systems for scientific applications.

Authors

Institutions

Publication Details

Journal
Cell Reports Physical Science
Published
2026-06-18
DOI
https://doi.org/10.1016/j.xcrp.2026.103400
Primary Topic
Adversarial Robustness in Machine Learning
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Scientific AI systems require domain-aware robustness validation

Haoning Dang, Fang Wang, Long-Gang Pang, Nan Cheng et al.
Cell Reports Physical Science
Adversarial Robustness in Machine Learning
article

Scientific AI systems require domain-aware robustness validation

Haoning Dang, Fang Wang, Long-Gang Pang, Nan Cheng, Shiyao Wang, Zehan Liu, Yiyan Zhang, J H Zhang, Deyu Meng, Wei Zhang, Duo Zhang, Fupeng Li, Jian-Nan Chen, Jiahao Song, Jianhui Xu, Chunxiang Shi, Xiu-Cheng Wang
article en

Abstract

AI has become an essential tool in many areas of scientific discovery. However, its robustness, particularly in high-stakes scientific applications, remains insufficiently understood. While adversarial vulnerabilities have been widely studied in image recognition, their implications in scientific disciplines, where physical constraints and numerical stability are critical, have not been thoroughly explored. This work presents a systematic evaluation of AI model fragility across five scientific domains: climate modeling, quantum chemistry, fluid dynamics, quantum chromodynamics, and wireless communications. We demonstrate that small, structured perturbations can destabilize predictions in all of these domains, even in models that achieve high accuracy under normal conditions. Furthermore, the failure patterns vary across domains, suggesting that these vulnerabilities are domain specific. Our findings underscore the importance of assessing AI robustness in scientific contexts and offer preliminary insights into how to design more resilient and reliable AI systems for scientific applications.

Cell Reports Physical ScienceVol. 7(7)
Tokyo Institute of Technology (JP), Xidian University (CN), Northwestern Polytechnical University (CN), Peking University (CN), Central China Normal University (CN), Microbiology Institute of Shaanxi (CN), Guangzhou Institute of Geography (CN), Northwest Institute of Nuclear Technology (CN), Xi'an Jiaotong University (CN), Shaanxi Normal University (CN)
Climate action
Openalex Percentile: Top 4%
Adversarial Robustness in Machine Learning
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.