Scholarlink

In the context of the regulation of so-called high-risk AI applications by the EU AI Act, the obligation to conduct risk management plays a decisive role. In theory, manufacturers and operators of these systems must already mitigate the risks posed by their systems during the development phase. However, this paper argues that there is a fundamental bias on the part of manufacturers and operators, which threatens to result in third-party risks in particular not being adequately taken into account. It is also shown that the concretization mechanisms of the relatively abstractly formulated AIA play a critical role in ensuring that third-party risks receive appropriate attention.

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

Fox Guarding the Chickens—Bias in Risk Management Obligations for High-Risk AI Systems under the EU AI Act

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Nils Brinker,
  • Richard Skalt

摘要

In the context of the regulation of so-called high-risk AI applications by the EU AI Act, the obligation to conduct risk management plays a decisive role. In theory, manufacturers and operators of these systems must already mitigate the risks posed by their systems during the development phase. However, this paper argues that there is a fundamental bias on the part of manufacturers and operators, which threatens to result in third-party risks in particular not being adequately taken into account. It is also shown that the concretization mechanisms of the relatively abstractly formulated AIA play a critical role in ensuring that third-party risks receive appropriate attention.