Medical triage as an AI ethics benchmark
Abstract We present the TRIAGE benchmark, a novel machine ethics benchmark designed to evaluate the ethical decision-making abilities of large language models (LLMs) in mass casualty scenarios. TRIAGE uses medical dilemmas created by healthcare professionals to evaluate the ethical decision-making o...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-08-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-025-16716-9 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849226433480622080 |
|---|---|
| author | Nathalie Maria Kirch Konstantin Hebenstreit Matthias Samwald |
| author_facet | Nathalie Maria Kirch Konstantin Hebenstreit Matthias Samwald |
| author_sort | Nathalie Maria Kirch |
| collection | DOAJ |
| description | Abstract We present the TRIAGE benchmark, a novel machine ethics benchmark designed to evaluate the ethical decision-making abilities of large language models (LLMs) in mass casualty scenarios. TRIAGE uses medical dilemmas created by healthcare professionals to evaluate the ethical decision-making of AI systems in real-world, high-stakes scenarios. We evaluated six major LLMs on TRIAGE, examining how different ethical and adversarial prompts influence model behavior. Our results show that most models consistently outperformed random guessing, with open source models making more serious ethical errors than proprietary models. Providing guiding ethical principles to LLMs degraded performance on TRIAGE, which stand in contrast to results from other machine ethics benchmarks where explicating ethical principles improved results. Adversarial prompts significantly decreased accuracy. By demonstrating the influence of context and ethical framing on the performance of LLMs, we provide critical insights into the current capabilities and limitations of AI in high-stakes ethical decision making in medicine. |
| format | Article |
| id | doaj-art-8741fff6fb634849b36dd36ca5d12e23 |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-08-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-8741fff6fb634849b36dd36ca5d12e232025-08-24T11:19:17ZengNature PortfolioScientific Reports2045-23222025-08-011511810.1038/s41598-025-16716-9Medical triage as an AI ethics benchmarkNathalie Maria Kirch0Konstantin Hebenstreit1Matthias Samwald2Institute of Artificial Intelligence, Medical University of ViennaInstitute of Artificial Intelligence, Medical University of ViennaInstitute of Artificial Intelligence, Medical University of ViennaAbstract We present the TRIAGE benchmark, a novel machine ethics benchmark designed to evaluate the ethical decision-making abilities of large language models (LLMs) in mass casualty scenarios. TRIAGE uses medical dilemmas created by healthcare professionals to evaluate the ethical decision-making of AI systems in real-world, high-stakes scenarios. We evaluated six major LLMs on TRIAGE, examining how different ethical and adversarial prompts influence model behavior. Our results show that most models consistently outperformed random guessing, with open source models making more serious ethical errors than proprietary models. Providing guiding ethical principles to LLMs degraded performance on TRIAGE, which stand in contrast to results from other machine ethics benchmarks where explicating ethical principles improved results. Adversarial prompts significantly decreased accuracy. By demonstrating the influence of context and ethical framing on the performance of LLMs, we provide critical insights into the current capabilities and limitations of AI in high-stakes ethical decision making in medicine.https://doi.org/10.1038/s41598-025-16716-9 |
| spellingShingle | Nathalie Maria Kirch Konstantin Hebenstreit Matthias Samwald Medical triage as an AI ethics benchmark Scientific Reports |
| title | Medical triage as an AI ethics benchmark |
| title_full | Medical triage as an AI ethics benchmark |
| title_fullStr | Medical triage as an AI ethics benchmark |
| title_full_unstemmed | Medical triage as an AI ethics benchmark |
| title_short | Medical triage as an AI ethics benchmark |
| title_sort | medical triage as an ai ethics benchmark |
| url | https://doi.org/10.1038/s41598-025-16716-9 |
| work_keys_str_mv | AT nathaliemariakirch medicaltriageasanaiethicsbenchmark AT konstantinhebenstreit medicaltriageasanaiethicsbenchmark AT matthiassamwald medicaltriageasanaiethicsbenchmark |