Sep 2026· Zenodo (CERN European Organization for Nuclear Research)
Occupational Health and Safety Research
Abstract
This article examines how aviation accidents can emerge from the progressive deterioration of organizational and institutional defenses rather than isolated human error. Drawing on the development of crew resource management, safety culture, quality assurance, voluntary reporting, and Safety Management Systems, the analysis explores how regulators, operators, manufacturers, military commands, training and maintenance organizations, airports, and investigative agencies form aviation’s multilayered safety architecture. Particular attention is given to normalization of deviance, staffing and experience pressures, weakened supervision, reporting culture, training deficiencies, organizational drift, and the distinction between regulatory compliance and operational resilience. Contemporary initiatives involving the Federal Aviation Administration, National Transportation Safety Board, manufacturers, the U.S. military, and DARPA are examined alongside the potential of artificial intelligence to identify institutional degradation before it enters an accident sequence. Legal and civil consequences are also considered when organizations recognize hazards but fail to act. The article concludes that the most dangerous safety system may be one that appears functional while progressively losing its ability to recognize and arrest deteriorating conditions. Future aviation safety depends upon detecting organizational drift, preserving institutional memory, protecting professional challenge, and using artificial intelligence to strengthen—not replace—the human and organizational defenses upon which aviation safety depends.
GAOKAO-Bench is introduced, an intuitive benchmark that employs questions from the Chinese GAOKAO examination as test samples, including both subjective and objective questions that contribute a robust evaluation benchmark for future large language models and offers valuable insights into the advantages and limitations of such models.
Xiaotian Zhang, Chun-yan Li, Yi Zong et al.· arXiv.org· 216 citations· ⚡17
This work investigates the possibilities of using LLMs in a resume screening setting via a document retrieval framework that simulates job candidate selection and finds that the MTEs are biased, significantly favoring White-associated names in 85% of cases and female-associated names in only 11.1% of cases.
Empirically, PRISM reduces the end-to-end time for data selection and model tuning to just 30% of conventional pipelines, and achieves this efficiency while simultaneously enhancing performance, surpassing models fine-tuned on the full dataset across eight multimodal and three language understanding benchmarks.
Jinhe Bi, Yifan Wang, Danqi Yan et al.· arXiv.org· 73 citations· ⚡4
The method, ECCOLA, is presented, which aims at making the high-level AI ethics principles more practical, making it possible for developers to more easily implement them in practice.
Ville Vakkuri, Kai-Kristian Kemell, P. Abrahamsson· EUROMICRO Conference on Soft...· 64 citations· ⚡6
This paper designs Markov decision processes (MDPs) for different combinatorial problems and proposes to train conditional GFlowNets to sample from the solution space and demonstrates that GFlowNet policies can efficiently find high-quality solutions.
Dinghuai Zhang, H. Dai, Esmeralda S. Whitammer et al.· Advances in Neural Informati...· 59 citations· ⚡8
An empirical study on the current state of practice in artificial intelligence ethics is conducted by means of a multiple case study of five case companies, which indicates a gap between research and practice in the area.
Ville Vakkuri, Kai-Kristian Kemell, Joni Kultanen et al.· arXiv.org· 56 citations· ⚡6