Ophelia Deroy
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Ophelia Deroy has a total of 96 co-authors. Here are the top co-authors by number of shared publications:
Charles Spence
Crossmodal Research Laboratory, Department of Experimental Psychology, University of Oxford, Oxford, UK
12 shared publications- Charles Spence
Department of Experimental Psychology, University of Oxford, Life and Mind Building, South Parks Road, Oxford OX1 3EL, UK
9 shared publications Bahador Bahrami
Munich Interactive Intelligence (Mi3), Ludwig Maximilian University (LMU), Munich, Germany; Crowd Cognition Group, Faculty of General Psychology and Education, LMU, Munich, Germany
7 shared publications- Bahador Bahrami
Intercultural Cognitive Network, Paris, France
5 shared publications - Merle Theresa Fairhurst
Centre for Tactile Internet with Human-in-the-Loop (CeTI), Faculty of Electrical and Computer Engineering, Technische Universität Dresden, Dresden, Germany
5 shared publications - Ana Tajadura-Jiménez
Universidad Carlos III de Madrid, Spain
4 shared publications
Latest Publications
- Open Access
AI is not as conscious as humans: Essentialism in mental state attributions to artificial systems
- Open Access
- 24 September 2026
- Published by Springer Nature in Philosophy & Technology
- Vol. 39 (4) , 189
Why We Need Epistemically, Not Morally, Trustworthy AI: Decision Support and the Need for Collective Accountability
Calls for “trustworthy AI” have become ubiquitous in policy and industry, yet the term remains conceptually underspecified. In earlier work (Dorsch & Deroy, 2024), we argued that moral trustworthiness is neither possible nor necessary for AI decision-support systems (AI-DSS), and indeed unethical to pursue, since it risks the category mistake of attributing benevolence to systems that lack it. This paper revisits that claim in light of recent critiques and developments, arguing that once moral trustworthiness is set aside, the central normative question shifts from whether AI systems can be trusted to how their epistemic influence on human decision-making ought to be justified and governed. We begin by situating our earlier thesis alongside complementary arguments that converge on the dangers of anthropomorphising AI. Second, we respond to critics who seek residual roles for moral trustworthiness and who clarify the epistemic conditions of trust. Third, we advance an account of epistemic trustworthiness as a property of AI outputs, rather than a property of the systems themselves. We contend that guaranteeing epistemic trustworthiness depends on human institutions, and is necessary to mitigate vulnerabilities introduced by integrating AI-DSS into hybrid decision-making. We connect this property to legal frameworks, showing how the EU AI Act’s ‘Right to Explanation’ codifies it as a necessary condition for deploying AI in high-risk domains.
- 24 September 2026
- Published by Springer Nature in Philosophy & Technology
- Vol. 39 (4) , 189
Open AccessWhy We Need Epistemically, Not Morally, Trustworthy AI: Decision Support and the Need for Collective Accountability
Calls for “trustworthy AI” have become ubiquitous in policy and industry, yet the term remains conceptually underspecified. In earlier work (Dorsch & Deroy, 2024), we argued that moral trustworthiness is neither possible nor necessary for AI decision-support systems (AI-DSS), and indeed unethical to pursue, since it risks the category mistake of attributing benevolence to systems that lack it. This paper revisits that claim in light of recent critiques and developments, arguing that once moral trustworthiness is set aside, the central normative question shifts from whether AI systems can be trusted to how their epistemic influence on human decision-making ought to be justified and governed. We begin by situating our earlier thesis alongside complementary arguments that converge on the dangers of anthropomorphising AI. Second, we respond to critics who seek residual roles for moral trustworthiness and who clarify the epistemic conditions of trust. Third, we advance an account of epistemic trustworthiness as a property of AI outputs, rather than a property of the systems themselves. We contend that guaranteeing epistemic trustworthiness depends on human institutions, and is necessary to mitigate vulnerabilities introduced by integrating AI-DSS into hybrid decision-making. We connect this property to legal frameworks, showing how the EU AI Act’s ‘Right to Explanation’ codifies it as a necessary condition for deploying AI in high-risk domains.
- Open Access
- 5 June 2026
- Published by Elsevier in La Presse Médicale
- Vol. 55 (4) , 104362
Collective intelligence in clinical medicine: what works, what fails, and how collaboration with AI really helps
- 5 June 2026
- Published by Elsevier in La Presse Médicale
- Vol. 55 (4) , 104362
Open AccessCollective intelligence in clinical medicine: what works, what fails, and how collaboration with AI really helps
- 27 March 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Peirce, le pragmatisme et les Grecs : dépendance à la réponse généralisée et réalisme
La thèse examine, à partir d’une relecture de Peirce et de certaines de ses interprétations des philosophes et des problèmes antiques, les arguments qui peuvent permettre à une conception pragmatiste des croyances et de la signification de parer aux accusations de conduire au relativisme. Ces arguments résident dans la façon dont ces conceptions s’articulent entre elles, et acceptent une forme de réalisme
- 27 March 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Peirce, le pragmatisme et les Grecs : dépendance à la réponse généralisée et réalisme
La thèse examine, à partir d’une relecture de Peirce et de certaines de ses interprétations des philosophes et des problèmes antiques, les arguments qui peuvent permettre à une conception pragmatiste des croyances et de la signification de parer aux accusations de conduire au relativisme. Ces arguments résident dans la façon dont ces conceptions s’articulent entre elles, et acceptent une forme de réalisme
- Open Access
- 24 March 2026
- Published by Walter de Gruyter GmbH in Multisensory Research
- Vol. 39 (3-5) , 399-415
When in Doubt, Touch Is More Convincing Than Vision
We sometimes feel compelled to touch objects after having looked at them. Is this because touch gives us more accurate information, or is it simply because we tend to trust what we feel more than what we see? To disentangle these possibilities, participants were subjected to a variant of the ‘Vertical–Horizontal’ illusion where the length of a vertical bar is consistently overestimated relative to the length of a horizontal bar. Our variant focuses on points of subjective equality where the bars almost look the same, and possesses two key characteristics: it creates ambiguity as stimuli can be subjectively similar yet objectively distinct and operates both in vision and touch. In a forced resampling bimodal experiment, participants inspected the stimuli using both vision and touch, judged which bar appeared longer, and rated their confidence. They then re-examined the same objects using either vision or touch, as instructed, and provided a second judgement with another confidence rating. Resampling did not significantly improve the overall accuracy, but participants were more likely to change their mind when using touch. In a second, free resampling bimodal experiment, participants chose their preferred modality for reinspection. Neither modality led to more accurate responses, but participants chose to resample by touch increasingly more frequently when the vertical and horizontal bars’ lengths were closer in perceived size, leading to higher ambiguity. Additionally, they were more likely to change their mind under these ambiguous conditions. These findings suggest a selective bias toward touch in case of ambiguity, even when it offers no objective advantage over vision.
- 24 March 2026
- Published by Walter de Gruyter GmbH in Multisensory Research
- Vol. 39 (3-5) , 399-415
Open AccessWhen in Doubt, Touch Is More Convincing Than Vision
We sometimes feel compelled to touch objects after having looked at them. Is this because touch gives us more accurate information, or is it simply because we tend to trust what we feel more than what we see? To disentangle these possibilities, participants were subjected to a variant of the ‘Vertical–Horizontal’ illusion where the length of a vertical bar is consistently overestimated relative to the length of a horizontal bar. Our variant focuses on points of subjective equality where the bars almost look the same, and possesses two key characteristics: it creates ambiguity as stimuli can be subjectively similar yet objectively distinct and operates both in vision and touch. In a forced resampling bimodal experiment, participants inspected the stimuli using both vision and touch, judged which bar appeared longer, and rated their confidence. They then re-examined the same objects using either vision or touch, as instructed, and provided a second judgement with another confidence rating. Resampling did not significantly improve the overall accuracy, but participants were more likely to change their mind when using touch. In a second, free resampling bimodal experiment, participants chose their preferred modality for reinspection. Neither modality led to more accurate responses, but participants chose to resample by touch increasingly more frequently when the vertical and horizontal bars’ lengths were closer in perceived size, leading to higher ambiguity. Additionally, they were more likely to change their mind under these ambiguous conditions. These findings suggest a selective bias toward touch in case of ambiguity, even when it offers no objective advantage over vision.
- Open Access
- 22 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 663-682
Mental Models in Human-AI Interaction: Systematic Review of Empirical Methodologies and Guidelines
- 22 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 663-682
Open AccessMental Models in Human-AI Interaction: Systematic Review of Empirical Methodologies and Guidelines
- Open Access
- 21 January 2026
- Published by Oxford University Press (OUP) in Social Cognitive and Affective Neuroscience
- Vol. 21 (1)
Covert neural and autonomic signatures of shared perception
Humans often co-perceive stimuli with others, yet the neurocognitive effects of such shared perceptual contexts are underexplored. We tested whether awareness that a visual stimulus is simultaneously available to another person, without interaction, modulates behavioral performance and neurophysiological signatures of perceptual decision-making. Thirty-three participants completed 640 trials of a Random Dot Kinematogram motion discrimination task while EEG and pupillometry data were recorded. A confederate was present, with a divider ensuring that, on each trial, the stimulus was either jointly visible to both or privately visible to the participant. Participants received no feedback and engaged in no interaction, isolating the effect of joint visibility. Behavioral performance was unaffected by social context, but EEG analysis revealed context-specific neural patterns emerging after cue onset and before stimulus presentation, suggesting proactive encoding of the social context. Additionally, pupil size was significantly greater during public visibility trials, indicating heightened arousal associated with social vigilance. These findings suggest that co-perception induces covert social vigilance—anticipatory arousal and neural readiness in response to co-visibility, even without interaction. Such covert markers could serve as biomarkers for altered social salience processing in clinical populations, such as those with social anxiety disorder or autism.
- 21 January 2026
- Published by Oxford University Press (OUP) in Social Cognitive and Affective Neuroscience
- Vol. 21 (1)
Open AccessCovert neural and autonomic signatures of shared perception
Humans often co-perceive stimuli with others, yet the neurocognitive effects of such shared perceptual contexts are underexplored. We tested whether awareness that a visual stimulus is simultaneously available to another person, without interaction, modulates behavioral performance and neurophysiological signatures of perceptual decision-making. Thirty-three participants completed 640 trials of a Random Dot Kinematogram motion discrimination task while EEG and pupillometry data were recorded. A confederate was present, with a divider ensuring that, on each trial, the stimulus was either jointly visible to both or privately visible to the participant. Participants received no feedback and engaged in no interaction, isolating the effect of joint visibility. Behavioral performance was unaffected by social context, but EEG analysis revealed context-specific neural patterns emerging after cue onset and before stimulus presentation, suggesting proactive encoding of the social context. Additionally, pupil size was significantly greater during public visibility trials, indicating heightened arousal associated with social vigilance. These findings suggest that co-perception induces covert social vigilance—anticipatory arousal and neural readiness in response to co-visibility, even without interaction. Such covert markers could serve as biomarkers for altered social salience processing in clinical populations, such as those with social anxiety disorder or autism.
- Open Access
The impact of labeling automotive AI as "trustworthy" or "reliable" on user evaluation and technology acceptance
This study explores whether labeling AI as "trustworthy" or "reliable" influences user perceptions and acceptance of automotive AI technologies. Using a one-way between-subjects design, the research involved 478 online participants who were presented with guidelines for either trustworthy or reliable AI. Participants then evaluated three vignette scenarios and completed a modified version of the Technology Acceptance Model, which included variables such as perceived ease of use, human-like trust, and overall attitude. Although labeling AI as "trustworthy" did not significantly influence judgments on specific scenarios, it increased perceived ease of use and human-like trust, particularly benevolence. This suggests a positive impact on usability and an anthropomorphic effect on user perceptions. The study provides insights into how specific labels can influence attitudes toward AI technology.
Open AccessThe impact of labeling automotive AI as "trustworthy" or "reliable" on user evaluation and technology acceptance
This study explores whether labeling AI as "trustworthy" or "reliable" influences user perceptions and acceptance of automotive AI technologies. Using a one-way between-subjects design, the research involved 478 online participants who were presented with guidelines for either trustworthy or reliable AI. Participants then evaluated three vignette scenarios and completed a modified version of the Technology Acceptance Model, which included variables such as perceived ease of use, human-like trust, and overall attitude. Although labeling AI as "trustworthy" did not significantly influence judgments on specific scenarios, it increased perceived ease of use and human-like trust, particularly benevolence. This suggests a positive impact on usability and an anthropomorphic effect on user perceptions. The study provides insights into how specific labels can influence attitudes toward AI technology.
- Open Access
- 17 July 2024
- Published by Wiley in Advanced Intelligent Systems
- Vol. 6 (10)
Shared Awareness Across Domain‐Specific Artificial Intelligence: An Alternative to Domain‐General Intelligence and Artificial Consciousness
Creating artificial general intelligence is the solution most often in the spotlight. It is also linked with the possibility—or fear—of machines gaining consciousness. Alternatively, developing domain‐specific artificial intelligence is more reliable, energy‐efficient, and ethically tractable, and raises mostly a problem of effective coordination between different systems and humans. Herein, it is argued that it will not require machines to be conscious and that simpler ways of sharing awareness are sufficient.
- 17 July 2024
- Published by Wiley in Advanced Intelligent Systems
- Vol. 6 (10)
Open AccessShared Awareness Across Domain‐Specific Artificial Intelligence: An Alternative to Domain‐General Intelligence and Artificial Consciousness
Creating artificial general intelligence is the solution most often in the spotlight. It is also linked with the possibility—or fear—of machines gaining consciousness. Alternatively, developing domain‐specific artificial intelligence is more reliable, energy‐efficient, and ethically tractable, and raises mostly a problem of effective coordination between different systems and humans. Herein, it is argued that it will not require machines to be conscious and that simpler ways of sharing awareness are sufficient.