Scholarlink

<p>Artificial intelligence (AI) is increasingly integrated in sustainability governance, yet most applications remain oriented towards optimisation and prediction, reducing complex social–ecological issues to technical problems. This narrow focus neglects plural values, lived experiences, and democratic judgement essential for transformative change. We advance a reflexive AI governance approach that treats AI as a socio-technical assemblage shaping problem framing, knowledge legitimisation, and authority distribution. Synthesising material, technical, epistemic, and ethico-political challenges, the paper draws on Aristotelian notions of <i>techne</i>, <i>episteme</i>, and <i>phronesis</i> to outline three reflexivity dimensions: design, epistemological, and engagement. Using a four-phase governance cycle and a protected area management scenario, we show how reflexivity can help align AI with plural, justice-oriented transformation pathways. Reflexive AI governance grounded in sustainability’s visions fosters deliberation, inclusivity, and ecological sufficiency, enabling democratic capacities over whether and how AI should be used, including the legitimate possibility of non-use, restriction, or withdrawal.</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

A reflexive artificial intelligence governance for transformative change in sustainability

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月29日

查看全文
  • Anna Hausmann,
  • Tuuli Toivonen,
  • Gonzalo Cortés-Capano

摘要

Artificial intelligence (AI) is increasingly integrated in sustainability governance, yet most applications remain oriented towards optimisation and prediction, reducing complex social–ecological issues to technical problems. This narrow focus neglects plural values, lived experiences, and democratic judgement essential for transformative change. We advance a reflexive AI governance approach that treats AI as a socio-technical assemblage shaping problem framing, knowledge legitimisation, and authority distribution. Synthesising material, technical, epistemic, and ethico-political challenges, the paper draws on Aristotelian notions of techne, episteme, and phronesis to outline three reflexivity dimensions: design, epistemological, and engagement. Using a four-phase governance cycle and a protected area management scenario, we show how reflexivity can help align AI with plural, justice-oriented transformation pathways. Reflexive AI governance grounded in sustainability’s visions fosters deliberation, inclusivity, and ecological sufficiency, enabling democratic capacities over whether and how AI should be used, including the legitimate possibility of non-use, restriction, or withdrawal.