Scholarlink

<p>Artificial intelligence systems shape how people argue and exchange reasons in digital and institutional contexts. This paper argues that AI systems act as co-arguers, influencing who is visible, credible, and heard in public reasoning. Drawing on argumentation theory and the framework of epistemic injustice, the study analyzes how algorithmic ranking, moderation, and summarization reshape the structure of argumentation. The analysis covers three domains social media, legal decision support, and educational feedback systems using qualitative case studies based on prior empirical research. The results show that algorithmic processes reinforce testimonial and hermeneutical exclusions by amplifying dominant viewpoints and limiting argumentative diversity. The paper proposes design and policy interventions, including participatory audits and diversity metrics, to reduce structural epistemic injustice. This approach reframes AI not as a neutral medium but as an active participant in shaping collective reasoning.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Algorithmic co-arguers: AI-mediated argumentation and structural epistemic injustice

  • 期刊论文

  • 发表日期：

    2026年3月30日

查看全文
  • Md Foysal Ahmed,
  • Nour-Elhouda Jraibi,
  • Md Salman Mahmud,
  • Adiba An Nur Oyshee,
  • Sazal Ahmed

摘要

Artificial intelligence systems shape how people argue and exchange reasons in digital and institutional contexts. This paper argues that AI systems act as co-arguers, influencing who is visible, credible, and heard in public reasoning. Drawing on argumentation theory and the framework of epistemic injustice, the study analyzes how algorithmic ranking, moderation, and summarization reshape the structure of argumentation. The analysis covers three domains social media, legal decision support, and educational feedback systems using qualitative case studies based on prior empirical research. The results show that algorithmic processes reinforce testimonial and hermeneutical exclusions by amplifying dominant viewpoints and limiting argumentative diversity. The paper proposes design and policy interventions, including participatory audits and diversity metrics, to reduce structural epistemic injustice. This approach reframes AI not as a neutral medium but as an active participant in shaping collective reasoning.