Adversarial Fragility and Language Vulnerability in Clinical AI: A Systematic Audit of Diagnostic Collapse Under Imperceptible Perturbations and Cross-Lingual Drift in Low-Resource Healthcare Settings
Fuente:
arXiv
Salvato in:
| Autori principali: | , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916018722439168 |
|---|---|
| author | Gabriel, Anthonio Oladimeji Yusuf, Ahmad Rufai |
| author_facet | Gabriel, Anthonio Oladimeji Yusuf, Ahmad Rufai |
| contents | Current clinical artificial intelligence (AI) systems are evaluated almost exclusively on clean, standardised, English-language inputs, conditions that do not reflect the realities of healthcare delivery in low-resource settings. This study presents the first systematic dual audit of two orthogonal safety vulnerabilities in clinical AI: adversarial image fragility and cross-lingual diagnostic drift. Using DenseNet121, the architecture underlying CheXNet, fine-tuned on the COVID-QU-Ex chest X-ray dataset (85,318 images; COVID-19, Non-COVID Pneumonia, Normal), we demonstrate that diagnostic accuracy collapses from 89.3% to 62.0% under a Fast Gradient Method (FGM) perturbation of epsilon=0.021, a magnitude imperceptible to the human eye. Standard defensive strategies including Gaussian smoothing and ensemble voting failed to restore clinical safety. In a parallel language fragility experiment, we tested Llama3.1:8b and NatLAS (N-ATLAS) on 20 COVID-19 clinical cases presented in Standard English, Nigerian Pidgin (Naija), and Yoruba-inflected English. Both models exhibited significant accuracy degradation: Llama3.1:8b dropped from 80.0% to 65.0% on Pidgin; NatLAS, an African-context model, collapsed from 85.0% to 55.0%, with diagnosis consistency falling to 50%. These findings establish a quantitative failure envelope for clinical AI under conditions representative of Primary Health Centre (PHC) deployment in Nigeria, and motivate urgent calls for adversarially hardened, linguistically inclusive clinical AI architectures. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_16993 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Adversarial Fragility and Language Vulnerability in Clinical AI: A Systematic Audit of Diagnostic Collapse Under Imperceptible Perturbations and Cross-Lingual Drift in Low-Resource Healthcare Settings Gabriel, Anthonio Oladimeji Yusuf, Ahmad Rufai Computers and Society Artificial Intelligence Machine Learning I.2.1; J.3 Current clinical artificial intelligence (AI) systems are evaluated almost exclusively on clean, standardised, English-language inputs, conditions that do not reflect the realities of healthcare delivery in low-resource settings. This study presents the first systematic dual audit of two orthogonal safety vulnerabilities in clinical AI: adversarial image fragility and cross-lingual diagnostic drift. Using DenseNet121, the architecture underlying CheXNet, fine-tuned on the COVID-QU-Ex chest X-ray dataset (85,318 images; COVID-19, Non-COVID Pneumonia, Normal), we demonstrate that diagnostic accuracy collapses from 89.3% to 62.0% under a Fast Gradient Method (FGM) perturbation of epsilon=0.021, a magnitude imperceptible to the human eye. Standard defensive strategies including Gaussian smoothing and ensemble voting failed to restore clinical safety. In a parallel language fragility experiment, we tested Llama3.1:8b and NatLAS (N-ATLAS) on 20 COVID-19 clinical cases presented in Standard English, Nigerian Pidgin (Naija), and Yoruba-inflected English. Both models exhibited significant accuracy degradation: Llama3.1:8b dropped from 80.0% to 65.0% on Pidgin; NatLAS, an African-context model, collapsed from 85.0% to 55.0%, with diagnosis consistency falling to 50%. These findings establish a quantitative failure envelope for clinical AI under conditions representative of Primary Health Centre (PHC) deployment in Nigeria, and motivate urgent calls for adversarially hardened, linguistically inclusive clinical AI architectures. |
| title | Adversarial Fragility and Language Vulnerability in Clinical AI: A Systematic Audit of Diagnostic Collapse Under Imperceptible Perturbations and Cross-Lingual Drift in Low-Resource Healthcare Settings |
| topic | Computers and Society Artificial Intelligence Machine Learning I.2.1; J.3 |
| url | https://arxiv.org/abs/2605.16993 |