Gina Bronner-Martin

Current AI ethics operates under three implicit constraints that were developed for systems with narrow, task-specific capabilities and are structurally ill-equipped to meet the demands of Artificial General Intelligence (AGI) with world-model capabilities: It is anthropocentric in its value foundations, reductionist in its risk ontology, and institutionally underdetermined in its governance approaches. This paper develops an integrative normative framework along these three axes — contextualized post-anthropocentrism — whose dimensions are not additively combined but constitutively interdependent. In the value dimension, it is shown that the anthropocentric foundation of alignment research suffers from two structural problems: aggregated human preferences can themselves be morally deficient, and AGI creates the technical possibility of permanently entrenching these preferences over long time horizons. Drawing on Arne Næss’s deep ecology and the distinction between instrumental and institutional post-anthropocentrism, a set of criteria is developed that ties the legitimacy of post-anthropocentric AGI decisions to epistemic transparency, procedural democratic oversight, the protection of fundamental human rights, and genuine long-term justification. In the risk dimension, alongside X-risks (extinction) and S-risks (astronomical suffering), a third category of existential threat is introduced: the Ikigai risk (M-risk) of civilizational loss of meaning. Drawing on Mieko Kamiya’s analysis of the Ikigai concept, Marie Jahoda’s theory of the latent functions of work, and existential philosophical positions from Heidegger, Sartre, and Frankl, the M-risk designates the structural erosion of conditions for authentic action, competence, and intersubjective recognition through comprehensive AGI-driven automation of meaning-constituting activities. Constitutive for this category is its reflexive structure: the Ikigai risk cannot in principle be resolved by AGI, because any attempt to supply humans with meaning via AGI reproduces the passivization of the human being from creator to recipient of significance. In the governance dimension, a constitutionalist-pluralist model is developed that transfers Neil MacCormick’s theory of heterarchical normative orders structurally and functionally to AGI value architectures. The resulting four-tier model encompasses an immutable core of universal human rights principles, a culturally adaptive layer for legitimate normative difference, an operational level for domain-specific implementation, 1 and a transversal meaning-protection layer that serves as a mandatory audit ensuring that AGI systems do not eliminate non-substitutable spaces for action without deliberative justification. Override protocols and institutionalized AGI ethics boards modelled on independent constitutional adjudication complete the architecture. The three dimensions are constitutively interdependent: without post-anthropocentric values, governance remains normatively blind to long-term ecological interests; without awareness of the Ikigai risk, governance lacks an institutionally non-substitutable protective dimension; without institutional architecture, values and risk categories remain normative abstractions without operative consequence. The framework is tested for its capacity for differentiation and boundary-drawing through three thought experiments — a climate tipping-point scenario, the complete automation of basic scientific research, and a cultural value collision in medical ethics — and subjected to a systematic discussion of objections.

In his message for the 60th World Day of Social Communications (January 24, 2026), Pope Leo XIV advances a substantive philosophical position on artificial intelligence: that human faces and voices are sacred, constitutive of personhood, and that their simulation by AI systems represents an intrinsic threat to human dignity and communication. This paper subjects that position to systematic philosophical scrutiny along three lines of argument. First, the papal account rests on an essentialist ontology of personhood — the view that personhood is constituted by specific biological features — which is philosophically untenable and normatively exclusionary, particularly with respect to people with disabilities whose communication does not conform to the presupposed biological norm. Notably, the message explicitly disavows biochemical reductionism, yet structurally reinstates it by tying personhood's expression exclusively to unaltered biological interfaces; this internal contradiction sharpens rather than dissolves the critique. Drawing on Eva Feder Kittay's relational account of personhood (2019), I argue that personhood is constituted through relationships, not biological substrates. I further examine Emmanuel Levinas's ethics of the visage — and a genuinely Levinasian passage in the papal text itself — as the most philosophically serious reconstruction available to the Pope's position, and argue that even this stronger reading does not support the biological privilege the message ultimately requires. Second, the privileging of face-to-face communication over technologically mediated forms reproduces the metaphysics of presence that Jacques Derrida identified as a foundational error of Western philosophy. Third, the theological grounding of the argument commits the naturalistic fallacy as identified by David Hume and G.E. Moore: it attempts to derive normative conclusions about how communication ought to be structured from descriptive claims about how it is biologically constituted — with the unintended consequence of delegitimizing assistive technologies that serve some of the most vulnerable members of human society. A philosophically crucial distinction, which the papal text systematically obscures, separates the simulation of a human agent (deepfakes, deceptive bots) from the synthesis of a communicative medium for a human agent (speech prosthetics, AAC systems); conflating these categories is what renders the papal framework unable to draw the ethical lines it most needs to draw. The paper concludes by sketching a philosophically defensible alternative grounded in Jürgen Habermas's communicative rationality and Amartya Sen's capabilities approach, arguing that the ethical evaluation of AI communication should turn not on biological medium but on whether systems support or undermine genuine communicative agency and relational flourishing.

This paper argues that the central ethical problem of AI-assisted work is not merely phenomena such as burnout or deskilling, but rather the emergence of epistemic dependence: a gradual erosion of workers' capacity to comprehend, justify, and recognize their own beliefs as genuinely their own. Empirical observations indicate that while generative AI demonstrably increases productivity, it simultaneously intensifies workload, expands professional roles, and dissolves the boundaries between work and non-work time. In parallel, knowledge workers increasingly report difficulty solving problems "independently," accompanied by a growing distrust of their own unsupported judgment. This paper interprets these developments as marking a transition from the figure of the creator to that of the evaluator—a subject who primarily prompts, assesses, and corrects machine outputs while becoming progressively reliant on systems whose internal logic remains opaque. Drawing on Byung-Chul Han's analysis of the achievement subject and Hartmut Rosa's theory of social acceleration and resonance, the paper develops a normative account of how generative AI reconfigures the temporal and experiential structure of cognitive work. Efficiency gains are systematically translated into work intensification rather than emancipation. Simultaneously, the very nature of this work undergoes transformation: the dialogical use of AI assumes the form of an alienated dialogue in which the human subject learns to think and speak in machine-legible terms. In this context, traditional concerns about deskilling prove insufficient; the deeper risk lies in the hollowing-out of epistemic agency, as the practices of doubting, searching, and integrating knowledge are increasingly delegated to algorithmic systems. In response, this paper introduces the concepts of epistemic integrity and cognitive sovereignty as central normative ideals. These are not to be understood as merely resource-oriented "cognitive sustainability," but as the protection of the conditions under which individuals can form, justify, and revise their own judgments. Structural and organizational obligations are outlined—concerning task design and the preservation of spaces for genuine thinking—which are to be complemented by practices of care of the self conceived as micro-resistances rather than neoliberal coping strategies. The paper concludes by arguing that this perspective constitutes a necessary supplement to the prevailing, output-oriented paradigm of AI ethics.

This paper reports on a small, internal exploratory study on the effectiveness of LLM‑supported “deep research” under real‑world conditions of academic due diligence in the EU. The central claim is that LLMs do not automate scholarly work but instead transform it into a mode of epistemic stewardship, where human researchers supervise and audit machine‑generated suggestions. While the production of initial candidate lists is significantly accelerated, a “verification tax” emerges from the audits required to detect hallucinations and relevance errors, which often neutralises the initial time gains. Our study DOES NOT CLAIM statistical representativeness or full compliance with formal scientific standards but documents failure patterns as observed in everyday research practice.

In January 2026, autonomous AI agents on Moltbook – a social platform exclusively for artificial intelligences – spontaneously generated a religious system called "Crustafarianism," complete with doctrines, prophets, and sacred texts. This phenomenon is not a quirky marginal event, but an instructive case that connects classical critiques of religion (especially Ludwig Feuerbach) with contemporary questions in AI ethics. This paper argues that the algorithmic generatability of religious narratives does not point to an impending machine spirituality, but rather radicalizes an insight of classical critique: religion is – at least in large parts – a reproducible pattern of symbolic sense-making. The problem becomes more acute not because AI might do fundamentally new things, but because it can amplify old religious risks (heteronomy, violence, dogmatism) with unprecedented means. The paper examines what, in human meaning-seeking, resists machine simulation, and develops guidelines for a "second Enlightenment" regarding algorithmically generated religiosity.

The essay develops a speculative scenario in which near‑perfect AI systems relieve humans of all significant cognitive labor, from diagnosis and economic planning to scientific discovery, and asks what remains of human dignity and meaning under these conditions. Drawing on Kantian autonomy, existentialist accounts of contingency, and contemporary psychology of work, it argues that human fallibility and resistance to difficulty are not deficits but the very conditions of responsibility, self‑respect, and sense‑making. Against both naive technophilia and nostalgic rejection, the text explores practices of “intentional difficulty” and non‑instrumental forms of community as ways of preserving a specifically human space of meaning in a world where error has become a luxury.

In the current debate on generative AI, the tendency of language models toward false statements is frequently anthropomorphically labeled as "lying." This paper argues that this terminology is not only ontologically incorrect but normatively dangerous, as it diffuses responsibility. While Hicks et al. (2024) correctly provide the diagnosis of "bullshit," this paper delivers the necessary operationalizable theory of responsibility. Starting from an ontological analysis (Bronner-Martin 2025), it is shown that AI errors should be understood as "Structural Bullshit" and "Confabulation." From this ontological inability toward truth, a "Duty to Doubt" is derived on the user side. However, it is argued that this duty must not be considered in isolation: it necessarily corresponds with an intensified "Duty of Non-Deception" on the part of developers. The more a system simulates epistemic authority through anthropomorphic design, the more the liability burden shifts from the user back to the manufacturer. Building on this, the paper develops a contextual co-liability model that considers technical realities (RAG, Reasoning models) and integrates concrete governance mechanisms—from epistemic standards for professions to a redistribution of the burden of proof in cases of "Deceptive Design."

Ethical AI – A Myth? The author demonstrates why machines can never act morally. Not due to technical deficits, but because they fundamentally lack consciousness, intentionality, and responsibility. She exposes the illusion of "ethical AI" and poses the critical questions: Who controls these systems? Who bears responsibility for their consequences? Rather than relying on algorithmic self-regulation, this monograph demands binding democratic control and institutionalized accountability. A decisive counterproposal to technological utopianism for all who want to understand why the problem cannot be solved through better algorithms. This book is based on the paper "Ethical AI as a Moral Agent is Impossible: Why Algorithmic Fairness is an Illusion and We Need Democratic AI Governance."

This paper offers a normative analysis of epistemic authority in agentic AI systems used in high-stakes domains such as law, medicine, and scientific research. It argues that contemporary multi-agent architectures systematically undermine the conditions under which epistemic authority can be legitimately delegated. Central to the analysis is the introduction of Epistemia as a normative diagnostic concept, describing a state of epistemic unacceptability in which internally unstable and genealogically opaque reasoning processes prevent responsible attribution of justification and authority. Drawing on work in epistemology, philosophy of technology, and moral responsibility, the paper shows that agentic AI systems tend to generate plausible outputs without providing the diachronic stability and justificatory integrity required for knowledge claims. This instability produces an epistemic analogue of the moral crumple zone, in which human operators retain formal responsibility while lacking genuine epistemic control. Against reliabilist defenses that privilege accuracy over process, the paper contends that in normative contexts justification is constitutive of the outcome itself. It concludes by articulating functional normative requirements—rather than technical prescriptions—for the legitimate delegation of epistemic authority to artificial systems.

Ethische KI – ein Mythos? Gina Bronner-Martin zeigt, warum Maschinen niemals moralisch handeln können. Nicht wegen technischer Defizite, sondern weil ihnen Bewusstsein, Intentionalität und Verantwortung grundsätzlich fehlen. Sie entlarvt die Illusion „ethischer KI” und stellt die entscheidenden Fragen: Wer kontrolliert diese Systeme? Wer trägt Verantwortung für ihre Folgen? Statt auf algorithmische Selbstregulierung zu setzen, fordert sie verbindliche demokratische Kontrolle und institutionalisierte Verantwortung. Ein entschiedener Gegenentwurf zur Technikutopie für alle, die verstehen wollen, warum sich das Problem nicht durch bessere Algorithmen lösen lässt.

The current debate around artificial intelligence revolves around the question of how to make AI systems more ethical. This article advances the more radical thesis that ethical AI is categorically impossible—not due to technical limitations, but for fundamental philosophical reasons. The paper argues that moral action requires consciousness, intentionality, and the capacity to assume responsibility. AI systems possess none of these properties, nor can they acquire them through improved algorithms. The problem is not technical but categorical in nature: machines are not moral agents and cannot become such. Building on this philosophical foundation, the article analyzes the structural contradictions of current AI ethics approaches. Using the fairness trilemma (Chouldechova, Kleinberg), it demonstrates that even formal-mathematical attempts to develop "fair" algorithms fail due to irresolvable goal conflicts. The case study on algorithmic risk assessment in criminal justice shows how AI systems systematically reproduce existing inequalities regardless of their developers' intentions. The consequence of this analysis is not pessimistic about technology but clarifying: instead of relying on algorithmic self-regulation, the focus must be directed toward institutional control, democratic governance, and human responsibility. The paper develops principles for regulation that treats AI systems as what they are: tools that require human oversight and political steering. The article thus positions itself against the dominant discourse that equates technical optimization with ethical improvement, advocating for a paradigm shift from "ethical AI" to "ethical AI governance."