Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.13268 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915739437367296 |
|---|---|
| author | Ghafoor, Zainab Islam, Md Shafiqul Howlader, Koushik Khondokar, Md Rasel Bhattacharjee, Tanusree Chakraborty, Sayantan Roy, Adrito Bhattacharjee, Ushashi Roy, Tirtho |
| author_facet | Ghafoor, Zainab Islam, Md Shafiqul Howlader, Koushik Khondokar, Md Rasel Bhattacharjee, Tanusree Chakraborty, Sayantan Roy, Adrito Bhattacharjee, Ushashi Roy, Tirtho |
| contents | Large Language Models (LLMs) are increasingly applied in healthcare, yet ensuring their ethical integrity and safety compliance remains a major barrier to clinical deployment. This work introduces a multi-agent refinement framework designed to enhance the safety and reliability of medical LLMs through structured, iterative alignment. Our system combines two generative models - DeepSeek R1 and Med-PaLM - with two evaluation agents, LLaMA 3.1 and Phi-4, which assess responses using the American Medical Association's (AMA) Principles of Medical Ethics and a five-tier Safety Risk Assessment (SRA-5) protocol. We evaluate performance across 900 clinically diverse queries spanning nine ethical domains, measuring convergence efficiency, ethical violation reduction, and domain-specific risk behavior. Results demonstrate that DeepSeek R1 achieves faster convergence (mean 2.34 vs. 2.67 iterations), while Med-PaLM shows superior handling of privacy-sensitive scenarios. The iterative multi-agent loop achieved an 89% reduction in ethical violations and a 92% risk downgrade rate, underscoring the effectiveness of our approach. This study presents a scalable, regulator-aligned, and cost-efficient paradigm for governing medical AI safety. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_13268 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Improving the Safety and Trustworthiness of Medical AI via Multi-Agent Evaluation Loops Ghafoor, Zainab Islam, Md Shafiqul Howlader, Koushik Khondokar, Md Rasel Bhattacharjee, Tanusree Chakraborty, Sayantan Roy, Adrito Bhattacharjee, Ushashi Roy, Tirtho Artificial Intelligence Large Language Models (LLMs) are increasingly applied in healthcare, yet ensuring their ethical integrity and safety compliance remains a major barrier to clinical deployment. This work introduces a multi-agent refinement framework designed to enhance the safety and reliability of medical LLMs through structured, iterative alignment. Our system combines two generative models - DeepSeek R1 and Med-PaLM - with two evaluation agents, LLaMA 3.1 and Phi-4, which assess responses using the American Medical Association's (AMA) Principles of Medical Ethics and a five-tier Safety Risk Assessment (SRA-5) protocol. We evaluate performance across 900 clinically diverse queries spanning nine ethical domains, measuring convergence efficiency, ethical violation reduction, and domain-specific risk behavior. Results demonstrate that DeepSeek R1 achieves faster convergence (mean 2.34 vs. 2.67 iterations), while Med-PaLM shows superior handling of privacy-sensitive scenarios. The iterative multi-agent loop achieved an 89% reduction in ethical violations and a 92% risk downgrade rate, underscoring the effectiveness of our approach. This study presents a scalable, regulator-aligned, and cost-efficient paradigm for governing medical AI safety. |
| title | Improving the Safety and Trustworthiness of Medical AI via Multi-Agent Evaluation Loops |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2601.13268 |