Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
Dual-Use Research and Publication Policies: A Comparison of Journals in Life Sciences and Artificial Intelligence
4
Zitationen
2
Autoren
2024
Jahr
Abstract
Introduction: Dual-use research (DUR) describes research with potential benefits that could be misapplied for harm. Policy on DUR is often limited to life sciences research. However, recently, there have been demonstrations of how research with dual-use concerns may extend beyond the life sciences to artificial intelligence (AI). One method of ensuring that research with dual-use concerns is not misapplied for harm is by censoring scientific journal articles. Journals may have policy on managing article submissions with potential DUR concerns. This study compared the policies of life science and AI journals toward DUR. Methods: Google Scholar Metrics and Scimago Journal and Country Rank were utilized to identify and select the top 20 publications in fields of life sciences and AI by specific metrics. The publicly accessible websites of each journal were searched to ascertain their publication policies regarding DUR. Journals and/or publishers were contacted if no policy was located. Results: From Google Scholar, 12/20 journals within the "Life Sciences & Earth Sciences" category had policies on DUR; from Scimago, 9/16 of the "Biochemistry, Genetics, and Molecular Biology" category had policies; and 8/19 of the "Immunology and Microbiology" category had policies. For AI journals, 2/13 journals from Google Scholar had policies; 4/15 journals from Scimago had policies. Conclusion: More journals in the life sciences have extant policies on how to handle article submissions with DUR concerns. Very few AI journals have policies.
Ähnliche Arbeiten
Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI
2019 · 8.561 Zit.
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
2019 · 8.452 Zit.
High-performance medicine: the convergence of human and artificial intelligence
2018 · 7.948 Zit.
BioBERT: a pre-trained biomedical language representation model for biomedical text mining
2019 · 6.797 Zit.
Proceedings of the 19th International Joint Conference on Artificial Intelligence
2005 · 5.781 Zit.