The Epistemic Cost of Opacity: How the Use of Artificial Intelligence Undermines the Knowledge of Medical Doctors in High-Stakes Contexts
Open accessThe DOI matches its Crossref record; 83 of the 83 references it cites were found in a registry or an index; 8 of the 46 sources its claims cite were checked for support.
Abstract
Abstract Artificial intelligent (AI) systems used in medicine are often very reliable and accurate, but at the price of their being increasingly opaque. This raises the question whether a system’s opacity undermines the ability of medical doctors to acquire knowledge on the basis of its outputs. We investigate this question by focusing on a case in which a patient’s risk of recurring breast cancer is predicted by an opaque AI system. We argue that, given the system’s opacity, as well as the possibility of malfunctioning AI systems, practitioners’ inability to check the correctness of their outputs, and the high stakes of such cases, the knowledge of medical practitioners is indeed undermined. They are lucky to form true beliefs based on the AI systems’ outputs, and knowledge is incompatible with luck. We supplement this claim with a specific version of the safety condition on knowledge, Safety*. We argue that, relative to the perspective of the medical doctor in our example case, his relevant beliefs could easily be false, and this despite his evidence that the AI system functions reliably. Assuming that Safety* is necessary for knowledge, the practitioner therefore doesn’t know. We address three objections to our proposal before turning to practical suggestions for improving the epistemic situation of medical doctors.
Overview
CoChat Summary
TL;DR
The paper argues that medical doctors who rely on opaque AI outputs in high-stakes contexts may form true beliefs but still lack knowledge because their beliefs fail a perspective-relative safety condition, Safety*. Using a hypothetical breast-cancer recurrence case, it claims opacity, correctness-blindness, potential malfunction, and high stakes make the doctor’s true belief lucky and therefore not knowledge.
Key findings
- In the “Cancer Risk” case, an opaque AI system predicts that patient P is not at risk—below a threshold of 0.05—and MD forms a true belief on that basis. The authors argue MD does not know because, relative to his perspective, the belief could easily be false if the system were malfunctioning; it is therefore not Safe*.
- Four features undermine MD’s knowledge: the system’s opacity, MD’s inability to check correctness directly, the real possibility of AI flaws/malfunctions/biases, and the high stakes for P and MD. MD’s evidence that the system works reliably does not remove the doubt that defeats knowledge.
- The paper blocks three objections: overgeneralization to perceptual expertise, expert testimony, CT scanners, and RCTs; undergeneralization from the breast-cancer case; and the claim that knowledge is practically irrelevant. It argues knowledge matters for moral/legal responsibility and informed consent.
- Practical remedies suggested are independent sources and “broken-leg” override, explainable AI or expert understanding of the system, and reliable validation/certification structures.
Methods
- Conceptual and philosophical analysis, not an empirical study. The paper uses a hypothetical medical case, “Cancer Risk,” plus the “Broken Clock” epistemological analogy.
- It develops and applies Safety*: “If S were to believe that p, then relative to her perspective, it could not easily be false that p,” drawing on Whiting.
- It addresses three objections and then proposes practical measures; the paper states there are no data, materials, or code to be made available.
Limitations
- The paper states no formal limitations section; its Data Availability statement says there are no data, materials, or code to be made available.
- Evident from the text: the argument rests on a hypothetical case and philosophical intuitions rather than empirical measurement, and its conclusion depends on accepting Safety* as necessary for knowledge and on pragmatic encroachment about high stakes.
- Evident from the text: the scope is limited to high-stakes, correctness-blind medical contexts; the authors acknowledge that complete understanding of AI inner workings remains improbable and that generalization depends on shared structural features.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
7 Valid · 5 Not valid · 19 Not checked · 58 Own results and interpretations
Every source the paper cites for this claim was checked and supports it.
In the paper
“Importantly, it is often not only the (global) inner workings of such systems that are hard to understand, but also the (local) causes of their individual outputs, for instance, why a certain CT scan was classified as depicting a tumor, or why a certain medical treatment was recommended for a patient”
What the sources say
- Supportedcites Reyes et al., 2020 source
The Introduction’s “Global versus Local” section distinguishes explanations of patterns driving a model’s predictions from explanations of why it made a specific prediction for a given input. The paper’s Essentials also says radiology AI systems can be difficult for humans to interpret, supporting the claim that both levels can be hard to understand. (LLM, conf=0.91)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“We can say that some AI systems are, to a certain extent, opaque. Sometimes the opacity of a system is due merely to the fact that the system is proprietary, but other times the reason is that its model involves sub-symbolic representations incomprehensible even to experts, or that it is too large or too complex to grasp”
What the sources say
- Not checked — no full text of the sourcecites Bathaee, 2018
- Partly supportedcites Ghassemi et al., 2021 sourceExperimental — accuracy not yet measured
The Introduction says that many modern AI models are too complex and high-dimensional to be easily understood, supporting opacity due to complexity. It does not identify proprietary status or sub-symbolic representations as reasons for opacity. (LLM, conf=0.88)
- Partly supportedcites Mann et al., 2023 sourceExperimental — accuracy not yet measured
Our copy supports most of the sentence: Foreign Representation (Architectural Opacity, p. 2) attributes opacity to AI systems using distributed, subsymbolic representations generated in training; Complexity (same section, p. 2) covers systems 'complex because of scale', 'especially severe for the millions of parameters of some neural networks'; and Intentional Concealment (Socio-Technical Opacity, p. 4) covers private developers seeking opacity to protect intellectual property or secure economic advantage, with 'relying on open-source software instead' as a remedy — the closest the paper comes to 'proprietary'. However, the added qualifier that these sub-symbolic representations are 'incomprehensible even to experts' is not the paper's position: it says representations are foreign 'to humans' while listing XAI strategies (saliency maps, network dissection, concept activation vectors, p. 3) to map them onto human concepts, and Section II (p. 1) explicitly holds opacity to be stakeholder-dependent ('a system can be transparent for an expert but opaque for a layperson'). The paper also never uses the term 'proprietary', treating that motive as one instance of a broader intentional-concealment category. (LLM, conf=0.78)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Assuming that Safety* is necessary for knowledge, the practitioner therefore doesn't know.”
plainSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Rests on an assumption not shown here.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We argue that, relative to the perspective of the medical doctor in our example case, his relevant beliefs could easily be false, and this despite his evidence that the AI system func -tions reliably.”
hedgedSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“So, given that it is a matter of luck that MD's belief is true, he doesn't know that P is not an at-risk patient.”
- Also stated in: § 2 The Intuitive Case for Undermined Knowledge · p. 7 · plain
“So his belief is true only by luck, and he does not have knowledge.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Rests on an assumption not shown here.
- Also stated in: § 2 The Intuitive Case for Undermined Knowledge · p. 7 · plain
37 more claims
Create a free account to see all 89 claims and to verify the 2 references still open.
- Probably not a substantial claim. 34
- It may state two separate claims as one. 19
Citations
Cites
High-performance medicine: the convergence of human and artificial intelligence
Topol, E. J. (2019) · Nature Medicine · doi:10.1038/s41591-018-0300-7 · Cited by 10,450 · cited 3×
“Also see Topol ( 2019 ).)”
Dissecting racial bias in an algorithm used to manage the health of populations
Obermeyer, Z., Powers, B., Vogeli, C., & Mullainathan, S. (2019) · Science · doi:10.1126/science.aax2342 · Cited by 7,137 · cited 1×
“( For an illustra - tive case, see Obermeyer et al., 2019 .)”
(2000) · Cited by 5,055
Is Justified True Belief Knowledge?
Gettier, E. L. (1963) · Analysis · doi:10.1093/analys/23.6.121 · Cited by 3,127 · cited 1×
“She argues that belief with such contents constitutes knowledge under the same conditions as belief with non-probabilistic contents, e.g. when the subject is not in a Gettier case or a fake barn case (Gettier, 1963 ; Goldman, 1976).”
Williamson, T. (2002) · doi:10.1093/019925656x.001.0001 · Cited by 2,788
Discrimination and Perceptual Knowledge
Goldman, A. I. (1976) · The Journal of Philosophy · doi:10.2307/2025679 · Cited by 1,870 · cited 1×
“She argues that belief with such contents constitutes knowledge under the same conditions as belief with non-probabilistic contents, e.g. when the subject is not in a Gettier case or a fake barn case (Gettier, 1963 ; Goldman, 1976).”
The false hope of current approaches to explainable artificial intelligence in health care
Ghassemi, M., Oakden-Rayner, L., & Beam, A. L. (2021) · The Lancet Digital Health · doi:10.1016/s2589-7500(21)00208-9 · Cited by 1,724 · cited 3×
“…of a system is due merely to the fact that the system is proprietary, but other times the reason is that its model involves sub-symbolic representations incomprehensible even to experts, or that it is too large or too complex to grasp (Bathaee, 2018 ; Ghassemi et al., 2021 ; Mann et al., 2023 ).”
Human-computer collaboration for skin cancer recognition
Tschandl, P., Rinner, C., Apalla, Z., Argenziano, G., Codella, N., Halpern, A., Janda, M., Lallas, A., Longo, C., Malvehy, J., et al. (2020) · Nature Medicine · doi:10.1038/s41591-020-0942-0 · Cited by 1,024 · cited 1×
“As examples, see the studies cited in Bjerring and Busch ( 2021 , 350) as well as Tschandl et al. ( 2020 ) or Tu et al. ( 2024 ).”
How to Defeat Opposition to Moore
Sosa, E. (1999) · Noûs · doi:10.1111/0029-4624.33.s13.7 · Cited by 969
Pritchard, D. (2005) · doi:10.1093/019928038x.001.0001 · Cited by 874
(2000) · Managing Media Services · doi:10.5040/9798216977261.0007 · Cited by 863
Nadarzynski, T., Miles, O., Cowie, A., & Ridge, D. (2019) · DIGITAL HEALTH · doi:10.1177/2055207619871808 · Cited by 810 · cited 1×
“Indeed, we think that it may be par - tially responsible for a certain skepticism and hesitancy towards AI in the healthcare domain (Hengstler et al., 2016 ; Nadarzynski et al., 2019 ; McCradden et al. 2020 ).”
Artificial Intelligence and Black‐Box Medical Decisions: Accuracy versus Explainability
London, A. J. (2019) · Hastings Center Report · doi:10.1002/hast.973 · Cited by 801 · cited 3×
“And similarly for medical doctors rely - ing on the expertise of their colleagues, on medical machinery such as CT scanners, or on the results of randomized controlled trials (RCTs) (London, 2019 ; Kempt et al., 2022 ) – see Table 1 .”
Evidence, Pragmatics, and Justification
Fantl, J., & McGrath, M. (2002) · The Philosophical Review · doi:10.1215/00318108-111-1-67 · Cited by 786 · cited 1×
“On the extreme end, such a worry has been articulated by philosophers as the claim that it is irrational to reason or to act on the basis of a fact unless you know that fact (Hawthorne & Stanley, 2008 ; similarly, Fantl & McGrath, 2002 ).”
Algorithmic bias: review, synthesis, and future research directions
Kordzadeh, N., & Ghasemaghaei, M. (2022) · European Journal of Information Systems · doi:10.1080/0960085x.2021.1927212 · Cited by 770 · cited 1×
“Further, it could be that because of some bias in the training data, the AI sys - tem regularly classifies a subgroup of the population with certain features as not at risk even if they are at risk (Fazelpour & Danks, 2021 ; Kordzadeh & Ghasemaghaei, 2021 ; Garcia, 2016 ).”
Hengstler, M., Enkel, E., & Duelli, S. (2016) · Technological Forecasting and Social Change · doi:10.1016/j.techfore.2015.12.014 · Cited by 661 · cited 1×
“Indeed, we think that it may be par - tially responsible for a certain skepticism and hesitancy towards AI in the healthcare domain (Hengstler et al., 2016 ; Nadarzynski et al., 2019 ; McCradden et al. 2020 ).”
Hawthorne, J., Stanley, J., & Journal of Philosophy, I. (2008) · Journal of Philosophy · doi:10.5840/jphil20081051022 · Cited by 657 · cited 1×
“On the extreme end, such a worry has been articulated by philosophers as the claim that it is irrational to reason or to act on the basis of a fact unless you know that fact (Hawthorne & Stanley, 2008 ; similarly, Fantl & McGrath, 2002 ).”
What do we need to build explainable AI systems for the medical domain?
Holzinger, A., Biemann, C., Pattichis, C. S., & Kell, D. B. (2017) · arXiv · doi:10.48550/arxiv.1712.09923 · Cited by 636 · cited 1×
“While a complete understanding of the inner workings remains improbable, approaching this with explainable AI may be valuable (Holz - inger, 2018 ; Holzinger et al., 2017).”
Epistemic Authority: A Theory of Trust, Authority, and Autonomy in Belief
Zagzebski, L. T. (2012) · doi:10.1093/acprof:oso/9780199936472.001.0001 · Cited by 489 · cited 1×
“Another difference is that, as some philosophers have argued, self-trust in one’s own capacities, together with the conscientious employment of these capacities, is an essential precondition for having justified beliefs and knowledge in the first place (Zagzebski, 2012 ; Dormandy, 2020 ).”
On the interpretability of artificial intelligence in radiology: challenges and opportunities
Reyes, M., Meier, R., Pereira, S., Silva, C. A., Dahlweid, F.-M., Tengg-Kobligk, H. V., Summers, R. M., & Wiest, R. (2020) · Radiology: Artificial Intelligence · doi:10.1148/ryai.2020190043 · Cited by 484 · cited 1×
“…not only the (global) inner workings of such systems that are hard to understand, but also the (local) causes of their individual outputs, for instance, why a certain CT scan was classified as depicting a tumor, or why a certain medical treatment was recommended for a patient (Reyes et al., 2020 ).”
Durán, J. M., & Jongsma, K. R. (2021) · Journal of Medical Ethics · doi:10.1136/medethics-2020-106820 · Cited by 446 · cited 1×
“A third way of overcoming epistemic problems discussed in the literature on medical AI involves rigorous validation of AI systems (Durán & Formanek, 2018 ; Durán & Jongsma, 2021 ; Topol, 2019 ; Ghassemi et al., 2021 ).”
The Inescapability of Gettier Problems
Zagzebski, L. (1994) · The Philosophical Quarterly · doi:10.2307/2220147 · Cited by 434 · cited 1×
“But as Pritchard ( 2005 , 2016 ) and Zagzebski ( 1994 ) have forcefully argued, knowledge is not compatible with luck.”
Moss, S. (2018) · Oxford Scholarship Online · doi:10.1093/oso/9780198792154.001.0001 · Cited by 427
Pritchard, D. (2007) · Synthese · doi:10.1007/s11229-006-9039-7 · Cited by 408 · cited 1×
“In recent decades, Safety has been the focus of much discussion (Pritchard, 2007; Sosa, 1999 ).”
NCCN Guidelines® Insights: Breast Cancer, Version 4.2021: Featured Updates to the NCCN Guidelines
Gradishar, W. J., Moran, M. S., Abraham, J., Aft, R., Agnese, D., Allison, K. H., Blair, S. L., Burstein, H. J., Dang, C., Elias, A. D., et al. (2021) · Journal of the National Comprehensive Cancer Network · doi:10.6004/jnccn.2021.0023 · Cited by 337
When shall we use our heads instead of the formula?
Meehl, P. E. (1957) · Journal of Counseling Psychology · doi:10.1037/h0047554 · Cited by 310 · cited 1×
“In identifiable “broken-leg” scenarios (Meehl, 1957), the physician needs to be able to override the AI system.”
From Machine Learning to Explainable AI
Holzinger, A. (2018) · 2018 World Symposium on Digital Intelligence for Systems and Machines (DISA) · doi:10.1109/disa.2018.8490530 · Cited by 306
(2018) · Harvard journal of law & technology · Cited by 295
Artificial Intelligence and Patient-Centered Decision-Making
Bjerring, J. C., & Busch, J. (2021) · Philosophy & Technology · doi:10.1007/s13347-019-00391-6 · Cited by 274 · cited 2×
“So, as Bjerring and Busch ( 2021 , 364) point out, there is indeed some information made available that can be used by medi - cal doctors to justify their belief that a certain treatment is effective.”
Algorithmic bias: Senses, sources, solutions
Fazelpour, S., & Danks, D. (2021) · Philosophy Compass · doi:10.1111/phc3.12760 · Cited by 272 · cited 1×
“Further, it could be that because of some bias in the training data, the AI sys - tem regularly classifies a subgroup of the population with certain features as not at risk even if they are at risk (Fazelpour & Danks, 2021 ; Kordzadeh & Ghasemaghaei, 2021 ; Garcia, 2016 ).”
Garcia, M. (2016) · World Policy Journal · doi:10.1215/07402775-3813015 · Cited by 257 · cited 3×
“But as Pritchard ( 2005 , 2016 ) and Zagzebski ( 1994 ) have forcefully argued, knowledge is not compatible with luck.”
Geographic Distribution of US Cohorts Used to Train Deep Learning Algorithms
Kaushal, A., Altman, R., & Langlotz, C. (2020) · JAMA · doi:10.1001/jama.2020.12067 · Cited by 239 · cited 1×
“Transfer of AI tools across different populations is a real-world problem, as evidenced by the studies in Kaushal et al. ( 2020 ), by AI systems used to detect intracranial hemorrhages (O’Connor, 2021 ), or by Watson for oncology, which is used world-wide despite being trained by doctors at one US…”
Understanding from Machine Learning Models
Sullivan, E. (2022) · The British Journal for the Philosophy of Science · doi:10.1093/bjps/axz035 · Cited by 221 · cited 1×
“…by several philosophers of science, who predominantly, but not exclusively, ask how the opacity of AI systems in scientific inquiry affects the ability of scientists to explain and understand natural phenomena (e.g. Sullivan, 2022 , Duede, 2023 , Boge, 2022 , Creel, 2020 z & Beisbart, 2024 ).”
Baldwin, D. R., Gustafson, J., Pickup, L., Arteta, C., Novotny, P., Declerck, J., Kadir, T., Figueiras, C., Sterba, A., Exell, A., et al. (2020) · Thorax · doi:10.1136/thoraxjnl-2019-214104 · Cited by 207 · cited 1×
“Some of the new AI methods are used to detect or classify what is there (e.g., pulmonary nodules) (Baldwin et al., 2020 ), others to predict how things will progress (e.g., a disease) (Dembrower et al., 2020 ), and still others to recommend certain behaviors (e.g., a treatment) (Liu et al., 2018 ).”
Transparency in Complex Computational Systems
Creel, K. A. (2020) · Philosophy of Science · doi:10.1086/709729 · Cited by 190 · cited 1×
“…by several philosophers of science, who predominantly, but not exclusively, ask how the opacity of AI systems in scientific inquiry affects the ability of scientists to explain and understand natural phenomena (e.g. Sullivan, 2022 , Duede, 2023 , Boge, 2022 , Creel, 2020 z & Beisbart, 2024 ).”
Kleinbart the Oblivious and Other Tales of Ignorance and Responsibility
Rosen, G., & Journal of Philosophy, I. (2008) · Journal of Philosophy · doi:10.5840/jphil20081051023 · Cited by 174 · cited 1×
“Both in the debate over moral responsibility and that over informed consent, there is room for discussion how extensive the required knowledge has to be (Keren & Lev, 2022 ) or whether knowledge is needed to ensure these goods (Rosen, 2008 ).”
Dembrower, K., Liu, Y., Azizpour, H., Eklund, M., Smith, K., Lindholm, P., & Strand, F. (2020) · Radiology · doi:10.1148/radiol.2019190872 · Cited by 163 · cited 1×
“Some of the new AI methods are used to detect or classify what is there (e.g., pulmonary nodules) (Baldwin et al., 2020 ), others to predict how things will progress (e.g., a disease) (Dembrower et al., 2020 ), and still others to recommend certain behaviors (e.g., a treatment) (Liu et al., 2018 ).”
Liu, C., Liu, X., Wu, F., Xie, M., Feng, Y., & Hu, C. (2018) · Journal of Medical Internet Research · doi:10.2196/11087 · Cited by 162 · cited 2×
“Some of the new AI methods are used to detect or classify what is there (e.g., pulmonary nodules) (Baldwin et al., 2020 ), others to predict how things will progress (e.g., a disease) (Dembrower et al., 2020 ), and still others to recommend certain behaviors (e.g., a treatment) (Liu et al., 2018 ).”
JUSTIFIED BELIEF IN A DIGITAL AGE: ON THE EPISTEMIC IMPLICATIONS OF SECRET INTERNET TECHNOLOGIES
Miller, B., & Record, I. (2013) · Episteme · doi:10.1017/epi.2013.11 · Cited by 153 · cited 1×
“Epistemic problems similar to ours have been discussed with respect to beliefs based on internet searches, social media, or other everyday software applications, where there also are opaque AI systems that provide users with outputs (Dahl, 2018 ; Grindrod, 2019 ; Miller & Record, 2013 ).”
Decision Making Criteria in Oncology
Glatzer, M., Panje, C. M., Sirén, C. A., Čihorić, N., & Putora, P. M. (2018) · Oncology · doi:10.1159/000492272 · Cited by 142
Self-citationGrounds for Trust: Essential Epistemic Opacity and Computational Reliabilism
Durán, J. M., & Formanek, N. (2018) · Minds and Machines · doi:10.1007/s11023-018-9481-6 · Cited by 138 · cited 1×
“A third way of overcoming epistemic problems discussed in the literature on medical AI involves rigorous validation of AI systems (Durán & Formanek, 2018 ; Durán & Jongsma, 2021 ; Topol, 2019 ; Ghassemi et al., 2021 ).”
Assertion and Isolated Second‐Hand Knowledge*
Lackey, J. (2011) · Assertion · doi:10.1093/acprof:oso/9780199573004.003.0011 · Cited by 130 · cited 1×
“…typically able to (and are expected to be able to) explain the reasons for their diagnosis or risk assessments to others, including medical doctors (Lackey, 2011 ): Medical doctors often jointly discuss their reasons for their judgments, and it is not uncommon to ask medical specialists for…”
Interpretable, not black-box, artificial intelligence should be used for embryo selection
Afnan, M. A. M., Liu, Y., Conitzer, V., Rudin, C., Mishra, A., Savulescu, J., & Afnan, M. (2021) · Human Reproduction Open · doi:10.1093/hropen/hoab040 · Cited by 112
Two Dimensions of Opacity and the Deep Learning Predicament
Boge, F. J. (2022) · Minds and Machines · doi:10.1007/s11023-021-09569-4 · Cited by 110
The Epistemic Condition of Moral Responsibility
Robichaud, P. (2023) · A Companion to Free Will · doi:10.1002/9781119210177.ch21 · Cited by 107
From Responsibility to Reason-Giving Explainable Artificial Intelligence
Baum, K., Mantel, S., Schmidt, E., & Speith, T. (2022) · Philosophy & Technology · doi:10.1007/s13347-022-00510-w · Cited by 103 · cited 1×
“If patients don’t receive the right treatment in such cases, we will thus not be able to hold medical doctors responsible – there is a responsibility gap (Baum et al., 2022 ).”
Self-citationMcCradden, M. D., Baba, A., Saha, A., Ahmad, S., Boparai, K., Fadaiefard, P., & Cusimano, M. D. (2020) · CMAJ Open · doi:10.9778/cmajo.20190151 · Cited by 101 · cited 1×
“Indeed, we think that it may be par - tially responsible for a certain skepticism and hesitancy towards AI in the healthcare domain (Hengstler et al., 2016 ; Nadarzynski et al., 2019 ; McCradden et al. 2020 ).”
Towards Conversational Diagnostic AI
Tu, T., Palepu, A., Schaekermann, M., Saab, K. K., Freyberg, J., Tanno, R., Wang, A., Li, B., Amin, M., Tomašev, N., et al. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2401.05654 · Cited by 99
Moss, S. J. (2013) · The Philosophical Review · doi:10.1215/00318108-1728705 · Cited by 99 · cited 2×
“As Moss shows, we do indeed naturally say things like, ‘she knows that her specimen most likely belongs to G hackmani ’ (cf. Moss, 2013 , p. 7) or ‘he knew that it was probably time to repaint the fence’.”
When Doctors and AI Interact: on Human Responsibility for Artificial Risks
Verdicchio, M., & Perin, A. (2022) · Philosophy & Technology · doi:10.1007/s13347-022-00506-6 · Cited by 90
Implementation of Artificial Intelligence Techniques for Cancer Detection
Patel, D., Shah, Y., Thakkar, N., Shah, K., & Shah, M. (2019) · Augmented Human Research · doi:10.1007/s41133-019-0024-3 · Cited by 87 · cited 1×
“For instance, AI-based applications are becoming more pro - ficient in predicting health-related outcomes such as diagnosing diseases like lung cancer (Patel et al., 2019 ) or predicting cancer recurrence for a specific patient (Van Booven et al., 2021 ).”
Savage, C. H., Tanwar, M., Elkassem, A. M. A., Sturdivant, A. B., Hamki, O., Sotoudeh, H., Sirineni, G. K. R., Singhal, A., Milner, D. M., Jones, J. G. A., et al. (2024) · American Journal of Roentgenology · doi:10.2214/ajr.24.31639 · Cited by 87
A Systematic Review of Artificial Intelligence in Prostate Cancer
Van Booven, D. J., Kuchakulla, M., Pai, R., Frech, F. S., Ramasahayam, R., Reddy, P., Parmar, M., Ramasamy, R., & Arora, H. (2021) · Research and Reports in Urology · doi:10.2147/rru.s268596 · Cited by 77 · cited 1×
“For instance, AI-based applications are becoming more pro - ficient in predicting health-related outcomes such as diagnosing diseases like lung cancer (Patel et al., 2019 ) or predicting cancer recurrence for a specific patient (Van Booven et al., 2021 ).”
Deep Learning Opacity in Scientific Discovery
Duede, E. (2023) · Philosophy of Science · doi:10.1017/psa.2023.8 · Cited by 74 · cited 1×
“…by several philosophers of science, who predominantly, but not exclusively, ask how the opacity of AI systems in scientific inquiry affects the ability of scientists to explain and understand natural phenomena (e.g. Sullivan, 2022 , Duede, 2023 , Boge, 2022 , Creel, 2020 z & Beisbart, 2024 ).”
The Importance of Understanding Deep Learning
Räz, T., & Beisbart, C. (2022) · Erkenntnis · doi:10.1007/s10670-022-00605-y · Cited by 69
Pritchard, D. (2004) · Journal of Philosophical Research · doi:10.5840/jpr_2004_18 · Cited by 65
Emotions as Social Information in Shared Decision-Making in Oncology
Treffers, T., & Putora, P. M. (2020) · Oncology · doi:10.1159/000505341 · Cited by 60 · cited 1×
“…is a multifaceted process with several factors relevant, such as the availability of treatment options (Panje et al. 2018 ), a wide range of decision-making criteria (Glatzer et al., 2020 ), and even emotions that may significantly impact the decision-making process (Treffers & Putora, 2020 ).”
Self-citationPark, Y. W., Choi, D., Park, J. E., Ahn, S. S., Kim, H., Chang, J. H., Kim, S. H., Kim, H. S., & Lee, S. (2021) · Scientific Reports · doi:10.1038/s41598-021-82467-y · Cited by 58
The Safety Condition for Knowledge
Rabinowitz, D. (2011) · Cited by 54
Lackey, J. (2006) · Philosophy Compass · doi:10.1111/j.1747-9991.2006.00035.x · Cited by 46 · cited 1×
“From an epistemological angle, according to non-reductionism about testimony, hearers have a defeasible right to believe what speakers in their community tell them by default, without positive reasons to trust the speakers (Lackey, 2006 ).”
Epistemic Self-Trust: It's Personal
Dormandy, K. (2024) · Episteme · doi:10.1017/epi.2020.49 · Cited by 42 · cited 1×
“Another difference is that, as some philosophers have argued, self-trust in one’s own capacities, together with the conscientious employment of these capacities, is an essential precondition for having justified beliefs and knowledge in the first place (Zagzebski, 2012 ; Dormandy, 2020 ).”
Appraising Black-Boxed Technology: the Positive Prospects
Dahl, E. S. (2018) · Philosophy & Technology · doi:10.1007/s13347-017-0275-1 · Cited by 41 · cited 1×
“Epistemic problems similar to ours have been discussed with respect to beliefs based on internet searches, social media, or other everyday software applications, where there also are opaque AI systems that provide users with outputs (Dahl, 2018 ; Grindrod, 2019 ; Miller & Record, 2013 ).”
Legal and Moral Responsibility
Duff, A. (2009) · Philosophy Compass · doi:10.1111/j.1747-9991.2009.00257.x · Cited by 40 · cited 2×
“…standard phrasing of this condition, an agent is responsible for an action only if she knows what she is doing; and she is excused and thus not responsible or blameworthy for an action if she is non-culpably ignorant of the negative consequences of her action (Duff, 2009 , p. 979; Wieland, 2017).”
Relative explainability and double standards in medical decision-making
Kempt, H., Heilinger, J., & Nagel, S. K. (2022) · Ethics and Information Technology · doi:10.1007/s10676-022-09646-x · Cited by 40 · cited 1×
“And similarly for medical doctors rely - ing on the expertise of their colleagues, on medical machinery such as CT scanners, or on the results of randomized controlled trials (RCTs) (London, 2019 ; Kempt et al., 2022 ) – see Table 1 .”
Neirotti, P., Pesce, D., & Battaglia, D. (2021) · Technological Forecasting and Social Change · doi:10.1016/j.techfore.2021.121088 · Cited by 39 · cited 1×
“The “broken-leg rule” pro - poses judgmental approaches can be more effective than AI systems for rare circum - stances, due to their capacity to take into consideration additional information that may be relevant (Neirotti et al., 2021 ).”
Challenges in evaluating the accuracy of AI-containing digital triage systems: A systematic review
Ilicki, J. (2022) · PLoS ONE · doi:10.1371/journal.pone.0279636 · Cited by 39
Knowledge, justification, and (a sort of) safe belief
Whiting, D. (2020) · Synthese · doi:10.1007/s11229-018-01905-z · Cited by 34 · cited 1×
“We will rely on a version of the principle that draws on Whiting ( 2020 ), which we call Safety*.”
Panje, C. M., Glatzer, M., Sirén, C. A., Plaßwilm, L., & Putora, P. M. (2018) · JCO Clinical Cancer Informatics · doi:10.1200/cci.18.00017 · Cited by 32 · cited 1×
“As a matter of fact, medical decision-making is a multifaceted process with several factors relevant, such as the availability of treatment options (Panje et al. 2018 ), a wide range of decision-making criteria (Glatzer et al., 2020 ), and even emotions that may significantly impact the…”
Self-citationArtificial intelligence for hepatitis evaluation
Liu, W., Liu, X., Peng, M., Chen, G.-Q., Liu, P.-H., Cui, X., Jiang, F., & Dietrich, C. F. (2021) · World Journal of Gastroenterology · doi:10.3748/wjg.v27.i34.5715 · Cited by 28
Pritchard, D. (2016) · Palgrave Macmillan UK eBooks · doi:10.1007/978-1-137-52692-2 · Cited by 24
Eyal, N. (2018) · Routledge eBooks · doi:10.4324/9781351028264-25 · Cited by 21
Laxar, D., Eitenberger, M., Maleczek, M., Kaider, A., Hammerle, F. P., & Kimberger, O. (2023) · BMC Medicine · doi:10.1186/s12916-023-03068-2 · Cited by 19
AI, Radical Ignorance, and the Institutional Approach to Consent
Steinberg, E. (2024) · Philosophy & Technology · doi:10.1007/s13347-024-00787-z · Cited by 11
Sources of Opacity in Computer Systems: Towards a Comprehensive Taxonomy
Mann, S., Crook, B., Kästner, L., Schomäcker, A., & Speith, T. (2023) · 2023 IEEE 31st International Requirements Engineering Conference Workshops (REW) · doi:10.1109/rew57809.2023.00063 · Cited by 10 · cited 1×
“…of a system is due merely to the fact that the system is proprietary, but other times the reason is that its model involves sub-symbolic representations incomprehensible even to experts, or that it is too large or too complex to grasp (Bathaee, 2018 ; Ghassemi et al., 2021 ; Mann et al., 2023 ).”
Editorial: Risks and Benefits of Adjuvants to Cancer Therapies
Dong, P., & Gewirtz, D. A. (2022) · Frontiers in Oncology · doi:10.3389/fonc.2022.913626 · Cited by 9
Informed Consent, Error and Suspending Ignorance: Providing Knowledge or Preventing Error?
Keren, A., & Lev, O. (2022) · Ethical Theory and Moral Practice · doi:10.1007/s10677-022-10277-w · Cited by 7 · cited 2×
“…to the standard view, a patient’s consent to a medical treatment is valid only if the patient knows – and even understands– relevant facts about her medical condition, the treatment options, possible side-effects of the available treat - ments, and the like (Eyal, 2019 , § 4.2, Keren & Lev, 2022 ).”
InfluentialGrindrod, J. (2025) · Inquiry · doi:10.1080/0020174x.2019.1688178 · Cited by 7 · cited 2×
“Epistemic problems similar to ours have been discussed with respect to beliefs based on internet searches, social media, or other everyday software applications, where there also are opaque AI systems that provide users with outputs (Dahl, 2018 ; Grindrod, 2019 ; Miller & Record, 2013 ).”
(2017) · Oxford University Press eBooks · doi:10.1093/oso/9780198779667.003.0017 · Cited by 2
Guzman-Alvarez, A. (2023) · D-Scholarship@Pitt (University of Pittsburgh) · Cited by 1
Sources of Opacity in Computer Systems: Towards a Comprehensive Taxonomy
Mann, S., Crook, B., Kästner, L., et al. (2023) · arXiv · doi:10.48550/arxiv.2307.14232
9. DISCRIMINATION AND PERCEPTUAL KNOWLEDGE
Goldman, A. I. (n.d.) · Causal Theories of Mind · doi:10.1515/9783110843828.174
6. IS JUSTIFIED TRUE BELIEF KNOWLEDGE?
Gettier, E. L. (n.d.) · Causal Theories of Mind · doi:10.1515/9783110843828.135
Evidence, Pragmatics, and Justification
Fantl, J., & McGrath, M. (2002) · The Philosophical Review · doi:10.2307/3182570
Cited by
“Does Black Box AI In Medicine Compromise Informed Consent?”
Director, S. (2025) · Philosophy & Technology · doi:10.1007/s13347-025-00860-1 · Cited by 25
Herrera, F., & Calderón, R. (2026) · Technology in Society · doi:10.1016/j.techsoc.2026.103302 · Cited by 36
The end of theory? AI and ignorance in financial markets
Svetlova, E., & Arnoldi, J. (2026) · Finance and Society · doi:10.1017/fas.2026.10032 · Cited by 2
“…in finance), and indeed also a literature on ignorance and AI, most of which is highlighting the opacity of AI systems (Emmert-Streib, Yli-Harja, and Dehmer, 2020; Boge, 2022; Borch and Min, 2022; Khan and Vice, 2022; Duede, 2023; Carloni, Berti, and Colantonio, 2025; Schmidt et al., 2025).”
Med-ViX-Ray: Enhancing explainable chest X-ray analysis with clinical knowledge graphs
Cieri, M., & Palomba, F. (2026) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2026.109313 · Cited by 2
Approaches to AI Integration in Education and Healthcare
Evangeline, S. I. (2025) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-5072-1.ch011 · Cited by 1
Metadata
- Authors
- Schmidt, E., Putora, P. M., & Fijten, R.
- Affiliations
- TU Dortmund University (DE)
- University of Bern (CH)
- Maastricht University Medical Centre (NL)
- Corresponding author
- Schmidt, Eva
- Venue
- Philosophy & Technology
- Published
- 2025-01-13
- Volume
- 38
- Issue
- 1
- ISSN
- 2210-5433 (ISSN-L)
- 2210-5441
- OpenAlex
- W4406334317
- Semantic Scholar
- 839345de6a83ea740bbb45af20641cf8916d0f6e
- Cited by
- 14
- References
- 77
- Citations by year
- 2026: 11 · 2025: 3
- FWCI
- 111.65
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0 (tdm)
- creativecommons.org/licenses/by/4.0 (vor)
- Locations
- Philosophy & Technology (publishedVersion, cc-by)
- Research Publications (Maastricht University) (publishedVersion, cc-by)
- Funders
- Technische Universität Dortmund
- Topics
- Social Sciences › Arts and Humanities › Philosophy › Epistemology, Ethics, and Metaphysics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Social Sciences › Decision Sciences › Information Systems and Management › Knowledge Management and Technology
- Keywords
- artificial intelligence in healthcare, algorithmic opacity, epistemic luck, breast cancer recurrence prediction, high-stakes decision making, trust in AI outputs, epistemic justification
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol and qwen3.8-omni-flash).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.