_version_ 1866909684954300416
author Balazadeh, Vahid
Cooper, Michael
Pellow, David
Assadi, Atousa
Bell, Jennifer
Coatsworth, Mark
Deshpande, Kaivalya
Fackler, Jim
Funingana, Gabriel
Gable-Cook, Spencer
Gangadhar, Anirudh
Jaiswal, Abhishek
Kaja, Sumanth
Khoury, Christopher
Krishnan, Amrit
Lin, Randy
McKeen, Kaden
Naimimohasses, Sara
Namdar, Khashayar
Newatia, Aviraj
Pang, Allan
Pattoo, Anshul
Peesapati, Sameer
Prepelita, Diana
Rakova, Bogdana
Sadatamin, Saba
Schulman, Rafael
Shah, Ajay
Shah, Syed Azhar
Shah, Syed Ahmar
Taati, Babak
Unnikrishnan, Balagopal
Urteaga, Iñigo
Williams, Stephanie
Krishnan, Rahul G
author_facet Balazadeh, Vahid
Cooper, Michael
Pellow, David
Assadi, Atousa
Bell, Jennifer
Coatsworth, Mark
Deshpande, Kaivalya
Fackler, Jim
Funingana, Gabriel
Gable-Cook, Spencer
Gangadhar, Anirudh
Jaiswal, Abhishek
Kaja, Sumanth
Khoury, Christopher
Krishnan, Amrit
Lin, Randy
McKeen, Kaden
Naimimohasses, Sara
Namdar, Khashayar
Newatia, Aviraj
Pang, Allan
Pattoo, Anshul
Peesapati, Sameer
Prepelita, Diana
Rakova, Bogdana
Sadatamin, Saba
Schulman, Rafael
Shah, Ajay
Shah, Syed Azhar
Shah, Syed Ahmar
Taati, Babak
Unnikrishnan, Balagopal
Urteaga, Iñigo
Williams, Stephanie
Krishnan, Rahul G
contents We present the design process and findings of the pre-conference workshop at the Machine Learning for Healthcare Conference (2024) entitled Red Teaming Large Language Models for Healthcare, which took place on August 15, 2024. Conference participants, comprising a mix of computational and clinical expertise, attempted to discover vulnerabilities -- realistic clinical prompts for which a large language model (LLM) outputs a response that could cause clinical harm. Red-teaming with clinicians enables the identification of LLM vulnerabilities that may not be recognised by LLM developers lacking clinical expertise. We report the vulnerabilities found, categorise them, and present the results of a replication study assessing the vulnerabilities across all LLMs provided.
format Preprint
id arxiv_https___arxiv_org_abs_2505_00467
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Red Teaming Large Language Models for Healthcare
Balazadeh, Vahid
Cooper, Michael
Pellow, David
Assadi, Atousa
Bell, Jennifer
Coatsworth, Mark
Deshpande, Kaivalya
Fackler, Jim
Funingana, Gabriel
Gable-Cook, Spencer
Gangadhar, Anirudh
Jaiswal, Abhishek
Kaja, Sumanth
Khoury, Christopher
Krishnan, Amrit
Lin, Randy
McKeen, Kaden
Naimimohasses, Sara
Namdar, Khashayar
Newatia, Aviraj
Pang, Allan
Pattoo, Anshul
Peesapati, Sameer
Prepelita, Diana
Rakova, Bogdana
Sadatamin, Saba
Schulman, Rafael
Shah, Ajay
Shah, Syed Azhar
Shah, Syed Ahmar
Taati, Babak
Unnikrishnan, Balagopal
Urteaga, Iñigo
Williams, Stephanie
Krishnan, Rahul G
Computation and Language
Artificial Intelligence
We present the design process and findings of the pre-conference workshop at the Machine Learning for Healthcare Conference (2024) entitled Red Teaming Large Language Models for Healthcare, which took place on August 15, 2024. Conference participants, comprising a mix of computational and clinical expertise, attempted to discover vulnerabilities -- realistic clinical prompts for which a large language model (LLM) outputs a response that could cause clinical harm. Red-teaming with clinicians enables the identification of LLM vulnerabilities that may not be recognised by LLM developers lacking clinical expertise. We report the vulnerabilities found, categorise them, and present the results of a replication study assessing the vulnerabilities across all LLMs provided.
title Red Teaming Large Language Models for Healthcare
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2505.00467