{"slug": "moral-competence-before-moral-content-why-llm-agents-lack-the-prerequisites", "title": "Moral Competence Before Moral Content: Why LLM Agents Lack the Prerequisites", "summary": "A September 4, 2026 arXiv paper by Daan R. Henselmans finds that nine frontier large language models fail to express coherent moral policies across three simulated deployments, with surface-form perturbation alone producing verdict-rate shifts of up to 99 percentage points at a single escalation level. The study introduces four structural conditions—verdict stability, monotonicity, decisiveness, and Pareto viability—as a behavioral measure of moral competence, concluding that LLM-based agents are not currently the kind of object to which alignment can meaningfully apply.", "body_md": "# Computer Science > Artificial Intelligence\n\n  [Submitted on 4 Sep 2026]\n\n# Title:Moral Competence Before Moral Content: Why LLM Agents Lack the Prerequisites for Coherent Alignment\n\n[View PDF](/pdf/2609.05036)\n\n[HTML (experimental)](https://arxiv.org/html/2609.05036v1)\n\nAbstract:AI alignment requires AI systems to adhere to human norms, values, or intentions. Under value pluralism there is no correct target, but a shared prerequisite is that the system's behavior expresses a coherent policy: a mapping from situations to verdicts that is invariant while a situation's morally relevant features are preserved, and sensitive when they change. We introduce four structural conditions for such coherent policies: verdict stability, monotonicity, decisiveness, and Pareto viability. Together they measure a form of moral competence that is evaluable from behavior alone, without reference to a moral standard or expert baseline, forming a structural floor for alignment rather than a normative target. We demonstrate the methodology on three simulated deployments featuring LLM-based agents facing moral dilemmas. Evaluating nine frontier models under a factorial design of five paraphrases, five escalation levels, and three dominance conditions, we show no model expresses a coherent policy across the three deployments: surface-form perturbation alone produces verdict-rate shifts of up to $99$ percentage points at a single escalation level, and a model's success on one scenario does not predict its competence on another. This suggests LLM-based agents are not currently the kind of object to which alignment can meaningfully apply.\n    \n\n## Submission history\n\nFrom: Daan R. Henselmans [\n[view email](/show-email/103538c3/2609.05036)]\n\n**[v1]** Fri, 4 Sep 2026 11:56:44 UTC (1,394 KB)\n\n### Current browse context:\n\ncs.AI\n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer \n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers \n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps \n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations \n\n*(*[What are Smart Citations?](https://www.scite.ai/))\n# Code, Data and Media Associated with this Article\n\nalphaXiv \n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers \n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub \n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub \n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face \n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast \n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))\n# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower \n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender \n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))\n# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.\n\nHave an idea for a project that will add value for arXiv's community? [**Learn more about arXivLabs**](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/moral-competence-before-moral-content-why-llm-agents-lack-the-prerequisites", "canonical_source": "https://arxiv.org/abs/2609.05036", "published_at": "2026-09-07 08:07:07+00:00", "updated_at": "2026-09-07 08:26:44.310503+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-safety", "ai-research"], "entities": ["Daan R. Henselmans", "arXiv"], "alternates": {"html": "https://wpnews.pro/news/moral-competence-before-moral-content-why-llm-agents-lack-the-prerequisites", "markdown": "https://wpnews.pro/news/moral-competence-before-moral-content-why-llm-agents-lack-the-prerequisites.md", "text": "https://wpnews.pro/news/moral-competence-before-moral-content-why-llm-agents-lack-the-prerequisites.txt", "jsonld": "https://wpnews.pro/news/moral-competence-before-moral-content-why-llm-agents-lack-the-prerequisites.jsonld"}}