Epistemic Trust as a Mechanism for Ethics Integration: Failure Modes and Design Principles from 70 Moral Imagination Workshops

• URL: http://arxiv.org/abs/2604.11281v1
• Date: Mon, 13 Apr 2026 10:39:10 GMT
• Title: Epistemic Trust as a Mechanism for Ethics Integration: Failure Modes and Design Principles from 70 Moral Imagination Workshops
• Abstract summary: Bottom-up responsible innovation initiatives seek to empower technology development teams to engage in ethical reflection.<n>Yet such initiatives frequently fail to achieve practitioner engagement.<n>This paper proposes a conceptual model linking intervention design to engagement outcomes.
• Score: 0.15686650621337614
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Bottom-up responsible innovation initiatives seek to empower technology development teams to engage in ethical reflection, yet such interventions frequently fail to achieve practitioner engagement. Why do some ethics interventions succeed while others are dismissed as irrelevant, adversarial, or disconnected from work? This paper proposes epistemic trust -- the degree to which practitioners regard an intervention, its facilitators, and its content as credible, relevant, and actionable -- as a conceptual model linking intervention design to engagement outcomes. Drawing on philosophical work on testimony and on practice-based qualitative analysis of over 70 moral imagination workshops with engineering teams between 2019 and 2025, we identify five dimensions of epistemic trust salient to ethics interventions (Relevance, Inclusivity, Agency, Authority, and Alignment) and present a typology of 23 failure modes that arise when these dimensions are inadequately addressed. We derive nine design principles for cultivating epistemic trust, grounded in our operationalisation of moral imagination through technomoral scenarios and structured deliberation. Our findings contribute to the literature on collaborative socio-technical integration by specifying conditions of uptake that existing frameworks leave undertheorised. We acknowledge limitations including selection effects from voluntary participation and the absence of formal outcome measures, and position our failure mode typology as practitioner hypotheses warranting further empirical validation.

Related papers

• The relationship between professional and general ethics in generative AI [5.775372519896796]
  General and professional ethics are increasingly conflicted in contemporary AI systems.<n>It is through this equilibrium that certain professional ethics are prioritized over others and implemented in practice.<n>We propose a formulation for how to intervene in and change AI models' favored ethics in professional practices.
  arXiv  Detail & Related papers  (2026-08-29T14:46:27Z)
• A game theory for foundation models shows new paths to rational cooperation through similarity inference [56.53805518668248]
  We show that foundation model agents engaging in optimal planning consistently converge to stable cooperation.<n>By shifting from decoupled to embedded agency, these agents model themselves as part of the universe they inhabit.
  arXiv  Detail & Related papers  (2026-08-04T17:24:33Z)
• Theoretical adversarial collaboration: a template [0.0]
  We propose a five-step template for conducting theoretical adversarial collaborations.<n>The procedure is designed to move disputants beyond familiar critique-reply-rejoinder exchanges.<n>The template was developed through reflection on, and implemented in, our own adversarial collaboration in philosophy and foundations of physics.
  arXiv  Detail & Related papers  (2026-07-17T15:17:37Z)
• Provably Secure Agent Guardrail [89.79561918065122]
  Existing defense architectures rely on empirical semantic guardrails and probabilistic large model adjudicators.<n>This paper proposes a new security paradigm for agents based on the fundamental limitations of logical reasoning.
  arXiv  Detail & Related papers  (2026-05-28T02:12:41Z)
• The Ethical Knowledge Gap: Dispersed Knowledge, Sensemaking Failures, and Epistemic Dependence [0.0]
  This paper argues that the persistent gap between ethical intention and ethical implementation is a structural condition.<n>Ethical software development remains stubbornly difficult despite two decades of normative frameworks, professional codes, and participatory methodologies.
  arXiv  Detail & Related papers  (2026-04-27T08:16:16Z)
• Large Language Models as a Semantic Interface and Ethical Mediator in Neuro-Digital Ecosystems: Conceptual Foundations and a Regulatory Imperative [83.99688944263843]
  This article introduces and substantiates the concept of Neuro-Linguistic Integration (NLI)<n>It is a novel paradigm for human-technology interaction where Large Language Models act as a key semantic interface between raw neural data and their social application.<n>We argue for the development of a second-order neuroethics,' focused not merely on neural data protection but on the ethics of AI-mediated semantic interpretation itself.
  arXiv  Detail & Related papers  (2026-03-18T07:34:17Z)
• Mirror: A Multi-Agent System for AI-Assisted Ethics Review [104.3684024153469]
  Mirror is an agentic framework for AI-assisted ethical review.<n>It integrates ethical reasoning, structured rule interpretation, and multi-agent deliberation within a unified architecture.
  arXiv  Detail & Related papers  (2026-02-09T03:38:55Z)
• AI Safety, Alignment, and Ethics (AI SAE) [0.0]
  Moral norms are adaptive mechanisms that render cooperation fitness-viable under selection pressure.<n>Current alignment approaches add ethics post hoc, treating it as an external constraint rather than embedding it as an evolutionary strategy for cooperation.<n>I propose a governance--embedding--representation pipeline linking moral representation learning to system-level design and institutional governance.
  arXiv  Detail & Related papers  (2025-09-28T20:52:36Z)
• The Cognitive Foundations of Economic Exchange: A Modular Framework Grounded in Behavioral Evidence [0.0]
  Evidence from early human societies suggests that reciprocity-not barter-was the foundational economic logic.<n>We introduce a minimal behavioral framework based on cognitive primitives--individual recognition, reciprocal credence, and cost-return sensitivity.<n>This framework offers a unified substrate for modeling trust, coordination, and economic behavior in both human and artificial systems.
  arXiv  Detail & Related papers  (2025-05-05T18:21:53Z)
• Disciplining Deliberation: A Sociotechnical Perspective on Machine Learning Trade-offs [0.0]
  Two prominent trade-offs in artificial intelligence are between predictive accuracy and fairness, and between predictive accuracy and interpretability.<n> prevailing interpretation views these formal trade-offs as directly corresponding to tensions between underlying social values.<n>I introduce a sociotechnical approach to examining the value implications of trade-offs.
  arXiv  Detail & Related papers  (2024-03-07T05:03:18Z)
• The Participatory Turn in AI Design: Theoretical Foundations and the Current State of Practice [64.29355073494125]
  This article aims to ground what we dub the "participatory turn" in AI design by synthesizing existing theoretical literature on participation. We articulate empirical findings concerning the current state of participatory practice in AI design based on an analysis of recently published research and semi-structured interviews with 12 AI researchers and practitioners.
  arXiv  Detail & Related papers  (2023-10-02T05:30:42Z)
• Towards CausalGPT: A Multi-Agent Approach for Faithful Knowledge Reasoning via Promoting Causal Consistency in LLMs [55.66353783572259]
  Causal-Consistency Chain-of-Thought harnesses multi-agent collaboration to bolster the faithfulness and causality of foundation models.<n>Our framework demonstrates significant superiority over state-of-the-art methods through extensive and comprehensive evaluations.
  arXiv  Detail & Related papers  (2023-08-23T04:59:21Z)
• Active Inference in Robotics and Artificial Agents: Survey and Challenges [51.29077770446286]
  We review the state-of-the-art theory and implementations of active inference for state-estimation, control, planning and learning. We showcase relevant experiments that illustrate its potential in terms of adaptation, generalization and robustness.
  arXiv  Detail & Related papers  (2021-12-03T12:10:26Z)
• Ethical Assurance: A practical approach to the responsible design, development, and deployment of data-driven technologies [0.0]
  Article offers contributions to the interdisciplinary project of responsible research and innovation in data science and AI. First, it provides a critical analysis of current efforts to establish practical mechanisms for algorithmic assessment. Second, it provides an accessible introduction to the methodology of argument-based assurance. Third, it establishes a novel version of argument-based assurance that we call 'ethical assurance'
  arXiv  Detail & Related papers  (2021-10-11T11:21:49Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.