Mitigating the impact of biased artificial intelligence in emergency decision-making
Name
s43856-022-00214-4.pdf
Description
Published version
Size
493.07 KB
Format
Adobe PDF
Checksum (MD5)
ec180cf38dd3857061000ac7ff5b81db
Author(s)
Adam, Hammaad
Balagopalan, Aparna
Alsentzer, Emily
Christia, Fotini
Ghassemi, Marzyeh
Date Issued
November 21, 2022
Journal
Communications Medicine
Publisher
Springer Science and Business Media LLC
Citation
Adam, Hammaad, Balagopalan, Aparna, Alsentzer, Emily, Christia, Fotini and Ghassemi, Marzyeh. 2022. "Mitigating the impact of biased artificial intelligence in emergency decision-making." Communications Medicine, 2 (1).
Version
Final published version
Abstract
Abstract
Background Prior research has shown that artificial intelligence (AI) systems often encode biases against minority subgroups. However, little work has focused on ways to mitigate the harm discriminatory algorithms can cause in high-stakes settings such as medicine. Methods In this study, we experimentally evaluated the impact biased AI recommendations have on emergency decisions, where participants respond to mental health crises by calling for either medical or police assistance. We recruited 438 clinicians and 516 non-experts to participate in our web-based experiment. We evaluated participant decision-making with and without advice from biased and unbiased AI systems. We also varied the style of the AI advice, framing it either as prescriptive recommendations or descriptive flags. Results Participant decisions are unbiased without AI advice. However, both clinicians and non-experts are influenced by prescriptive recommendations from a biased algorithm, choosing police help more often in emergencies involving African-American or Muslim men. Crucially, using descriptive flags rather than prescriptive recommendations allows respondents to retain their original, unbiased decision-making. Conclusions Our work demonstrates the practical danger of using biased models in health contexts, and suggests that appropriately framing decision support can mitigate the effects of AI bias. These findings must be carefully considered in the many real-world clinical scenarios where inaccurate or biased models may be used to inform important decisions.
Background Prior research has shown that artificial intelligence (AI) systems often encode biases against minority subgroups. However, little work has focused on ways to mitigate the harm discriminatory algorithms can cause in high-stakes settings such as medicine. Methods In this study, we experimentally evaluated the impact biased AI recommendations have on emergency decisions, where participants respond to mental health crises by calling for either medical or police assistance. We recruited 438 clinicians and 516 non-experts to participate in our web-based experiment. We evaluated participant decision-making with and without advice from biased and unbiased AI systems. We also varied the style of the AI advice, framing it either as prescriptive recommendations or descriptive flags. Results Participant decisions are unbiased without AI advice. However, both clinicians and non-experts are influenced by prescriptive recommendations from a biased algorithm, choosing police help more often in emergencies involving African-American or Muslim men. Crucially, using descriptive flags rather than prescriptive recommendations allows respondents to retain their original, unbiased decision-making. Conclusions Our work demonstrates the practical danger of using biased models in health contexts, and suggests that appropriately framing decision support can mitigate the effects of AI bias. These findings must be carefully considered in the many real-world clinical scenarios where inaccurate or biased models may be used to inform important decisions.
MIT Department
Massachusetts Institute of Technology. Institute for Data, Systems, and Society
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Harvard University--MIT Division of Health Sciences and Technology
Massachusetts Institute of Technology. Institute for Medical Engineering & Science
Massachusetts Institute of Technology. Sociotechnical Systems Research Center
Massachusetts Institute of Technology. Department of Political Science
Terms of Use
Creative Commons Attribution 4.0 International license
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1038/s43856-022-00214-4
