Scholarlink

<p>The impact of Artificial Intelligence (AI) on clinical decision-making has become a focal point within a multitude of ethical, legal, and social discussions. Examining concrete use cases reveals that the introduction of an AI-system as a new actant in clinical practice changes the whole context with its relationships, roles, and routines. To better understand these changes, an interpretative framework can help to trace such transformations and uncover their ethical implications. In Science and Technology Studies (STS), a well-known approach is the application of Actor-Network-Theory (ANT) to concrete phenomena, including in the field of medicine. We aim to build on this established approach by developing a normative framework informed by ANT. To this end, we derive evaluation questions from Latour’s four meanings of mediation: <i>translation</i> to ask what goals should be pursued, <i>composition</i> to evaluate who should be involved, <i>reversible blackboxing</i> to specify how the different actants should be involved, and <i>delegation</i> to reflect on who should be responsible. These questions can thus inform the governance of AI-systems in medicine and contribute to their responsible implementation.</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

An actor-network-theory informed normative framework for the governance of AI-systems in medicine

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月15日

查看全文
  • Michael Hahn,
  • David Samhammer,
  • Max Tretter,
  • Peter Dabrock

摘要

The impact of Artificial Intelligence (AI) on clinical decision-making has become a focal point within a multitude of ethical, legal, and social discussions. Examining concrete use cases reveals that the introduction of an AI-system as a new actant in clinical practice changes the whole context with its relationships, roles, and routines. To better understand these changes, an interpretative framework can help to trace such transformations and uncover their ethical implications. In Science and Technology Studies (STS), a well-known approach is the application of Actor-Network-Theory (ANT) to concrete phenomena, including in the field of medicine. We aim to build on this established approach by developing a normative framework informed by ANT. To this end, we derive evaluation questions from Latour’s four meanings of mediation: translation to ask what goals should be pursued, composition to evaluate who should be involved, reversible blackboxing to specify how the different actants should be involved, and delegation to reflect on who should be responsible. These questions can thus inform the governance of AI-systems in medicine and contribute to their responsible implementation.