Who Decides What Is Harmful? Content Moderation Policy Through A Multi-Agent Personalised Inference Framework
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866909010242830336 |
|---|---|
| author | Gajewska, Ewelina Wawer, Michal Budzynska, Katarzyna Chudziak, Jaroslaw A. |
| author_facet | Gajewska, Ewelina Wawer, Michal Budzynska, Katarzyna Chudziak, Jaroslaw A. |
| contents | The increasing scale and complexity of online platforms raises critical policy questions around harmful content, digital well-being, and user autonomy. Traditional content moderation systems rely on centralised, top-down rules, often failing to accommodate the subjective nature of harm perception. This paper proposes an LLM-based multi-agent personalised inference framework that filters content based on unique sensitivity profiles of individual users. Our architecture combines domain-specific Expert Agents, a Manager Agent for orchestrating content analysis and agent selection, and a Ghost Profile Agent for simulating user perspectives, to inform moderation decisions. Evaluated against a range of non-personalised baselines, the system demonstrates up to a 32% improvement in accuracy, showing increased alignment with individual user sensitivities. Beyond technical performance, our framework provides policy-relevant insights for platform governance, providing a scalable way to reconcile moderation policies with societal and individual digital rights |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_01416 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Who Decides What Is Harmful? Content Moderation Policy Through A Multi-Agent Personalised Inference Framework Gajewska, Ewelina Wawer, Michal Budzynska, Katarzyna Chudziak, Jaroslaw A. Computers and Society Computation and Language The increasing scale and complexity of online platforms raises critical policy questions around harmful content, digital well-being, and user autonomy. Traditional content moderation systems rely on centralised, top-down rules, often failing to accommodate the subjective nature of harm perception. This paper proposes an LLM-based multi-agent personalised inference framework that filters content based on unique sensitivity profiles of individual users. Our architecture combines domain-specific Expert Agents, a Manager Agent for orchestrating content analysis and agent selection, and a Ghost Profile Agent for simulating user perspectives, to inform moderation decisions. Evaluated against a range of non-personalised baselines, the system demonstrates up to a 32% improvement in accuracy, showing increased alignment with individual user sensitivities. Beyond technical performance, our framework provides policy-relevant insights for platform governance, providing a scalable way to reconcile moderation policies with societal and individual digital rights |
| title | Who Decides What Is Harmful? Content Moderation Policy Through A Multi-Agent Personalised Inference Framework |
| topic | Computers and Society Computation and Language |
| url | https://arxiv.org/abs/2605.01416 |