# MedFailBench: A Clinician-Built Open-Source Benchmark for Medical AI Safety Boundary Inspection

arXiv: https://arxiv.org/abs/2607.15166 · HTML: https://mlsanity.com/paper/2607.15166/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 3 of its 20 references that are in the map.
- First posted on arXiv 16 Jul 2026.

## Builds on (most cited first, 3 of 3)

- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Large language models provide unsafe answers to patient-posed medical questions](https://mlsanity.com/paper/2507.18905/) (2025)
- [Setting Standards in Turkish NLP: TR-MMLU for Large Language Model Evaluation](https://mlsanity.com/paper/2501.00593/) (2024)

## Abstract (arXiv)

Most medical AI benchmarks measure whether a model knows the correct answer. MedFailBench asks a different question: which safety boundary failed? We present a synthetic benchmark and failure atlas built by a clinician. The resource labels medical AI errors by severity from 1 to 5 and safety gate type: missed urgent escalation, unsafe remote dosing, unsafe discharge reassurance, evidence fabrication, unsafe protocol execution, and source support gap. The current public release (v0.2.1) contains 44 synthetic cases reviewed by a clinician, with severity annotations, a public Hugging Face Space source, a safety gate taxonomy, a clinical severity rubric, and an automated pipeline for archiving model response screening runs. Forty cases have a populated safety gate field, and four require gate completion. No patient data, clinical validation claims, or model rankings are included. MedFailBench is released under Apache-2.0 and CC-BY-4.0 and carries the Zenodo DOI 10.5281/zenodo.21205535.