When Is It Safe to Introduce an AI System Into Healthcare? A Practical Decision Algorithm for the Ethical Implementation of Black-Box AI in Medicine
There is mounting global interest in the revolutionary potential of AI tools. However, its use in healthcare carries certain risks. Some argue that opaque (‘black box’) AI systems in particular undermine patients' informed consent. While interpretable models offer an alternative, this approach...
| Authors: | ; ; |
|---|---|
| Format: | Electronic Article |
| Language: | English |
| Check availability: | HBZ Gateway |
| Interlibrary Loan: | Interlibrary Loan for the Fachinformationsdienste (Specialized Information Services in Germany) |
| Published: |
2026
|
| In: |
Bioethics
Year: 2026, Volume: 40, Issue: 1, Pages: 61-72 |
| Further subjects: | B
black-box AI
B Informed Consent B Risk assessment B Artificial Intelligence B Clinical Practice B large language models |
| Online Access: |
Volltext (kostenfrei) Volltext (kostenfrei) |
MARC
| LEADER | 00000naa a22000002c 4500 | ||
|---|---|---|---|
| 001 | 1946468142 | ||
| 003 | DE-627 | ||
| 005 | 20251218101915.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 251218s2026 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1111/bioe.70032 |2 doi | |
| 035 | |a (DE-627)1946468142 | ||
| 035 | |a (DE-599)KXP1946468142 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rda | ||
| 041 | |a eng | ||
| 084 | |a 1 |2 ssgn | ||
| 100 | 1 | |a Allen, Jemima Winifred |e VerfasserIn |0 (orcid)0000-0002-3950-9950 |4 aut | |
| 245 | 1 | 0 | |a When Is It Safe to Introduce an AI System Into Healthcare? A Practical Decision Algorithm for the Ethical Implementation of Black-Box AI in Medicine |
| 264 | 1 | |c 2026 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a Computermedien |b c |2 rdamedia | ||
| 338 | |a Online-Ressource |b cr |2 rdacarrier | ||
| 520 | |a There is mounting global interest in the revolutionary potential of AI tools. However, its use in healthcare carries certain risks. Some argue that opaque (‘black box’) AI systems in particular undermine patients' informed consent. While interpretable models offer an alternative, this approach may be impossible with generative AI and large language models (LLMs). Thus, we propose that AI tools should be evaluated for clinical use based on their implementation risk, rather than interpretability. We introduce a practical decision algorithm for the clinical implementation of black-box AI by evaluating its risk of implementation. Applied to the case of an LLM for surgical informed consent, we assess a system's implementation risk by evaluating: (1) technical robustness, (2) implementation feasibility and (3) analysis of harms and benefits. Accordingly, the system is categorised as minimal-risk (standard use), moderate-risk (innovative use) or high-risk (experimental use). Recommendations for implementation are proportional to risk, requiring more oversight for higher-risk categories. The algorithm also considers the system's cost-effectiveness and patients' informed consent. | ||
| 601 | |a Implementation | ||
| 650 | 4 | |a Risk assessment | |
| 650 | 4 | |a large language models | |
| 650 | 4 | |a Informed Consent | |
| 650 | 4 | |a Clinical Practice | |
| 650 | 4 | |a black-box AI | |
| 650 | 4 | |a Artificial Intelligence | |
| 700 | 1 | |e VerfasserIn |0 (DE-588)1035485133 |0 (DE-627)747586047 |0 (DE-576)383073898 |4 aut |a Wilkinson, Dominic | |
| 700 | 1 | |a Savulescu, Julian |e VerfasserIn |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t Bioethics |d Oxford [u.a.] : Wiley-Blackwell, 1987 |g 40(2026), 1, Seite 61-72 |h Online-Ressource |w (DE-627)271596708 |w (DE-600)1480658-7 |w (DE-576)078707986 |x 1467-8519 |7 nnas |
| 773 | 1 | 8 | |g volume:40 |g year:2026 |g number:1 |g pages:61-72 |
| 856 | 4 | 0 | |u https://doi.org/10.1111/bioe.70032 |x Resolving-System |z kostenfrei |3 Volltext |7 0 |
| 856 | 4 | 0 | |u https://onlinelibrary.wiley.com/doi/abs/10.1111/bioe.70032 |x Verlag |z kostenfrei |3 Volltext |7 0 |
| 912 | |a NOMM | ||
| 935 | |a mteo | ||
| 951 | |a AR | ||
| ELC | |a 1 | ||
| ITA | |a 1 |t 1 | ||
| LOK | |0 000 xxxxxcx a22 zn 4500 | ||
| LOK | |0 001 4829804831 | ||
| LOK | |0 003 DE-627 | ||
| LOK | |0 004 1946468142 | ||
| LOK | |0 005 20251218055508 | ||
| LOK | |0 008 251218||||||||||||||||ger||||||| | ||
| LOK | |0 035 |a (DE-Tue135)IxTheo#2025-12-17#1F287F79D6C87C1429A7823C6C33C33EE9215A9A | ||
| LOK | |0 040 |a DE-Tue135 |c DE-627 |d DE-Tue135 | ||
| LOK | |0 092 |o n | ||
| LOK | |0 852 |a DE-Tue135 | ||
| LOK | |0 852 1 |9 00 | ||
| LOK | |0 935 |a ixzs |a zota | ||
| LOK | |0 939 |a 18-12-25 |b l01 | ||
| OAS | |a 1 | ||
| ORI | |a TA-MARC-ixtheoa001.raw | ||
