Scholarlink

This chapter reframes AI security as a systemic and socio-technical challenge rather than a purely technical task. Moving beyond traditional notions of access control and component-level protection, it introduces resilience, context-aware governance, and multi-layered defense as foundational principles for securing AI systems. The concept of systemic assurance is presented as a framework for identifying emergent vulnerabilities arising from interdependencies, distributional drift, adversarial adaptation, and feedback loops in real-world deployments.

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

Systemic Security, Resilience, and Multi-level Governance in AI

  • 章节

  • 发表日期：

    2026年

查看全文
  • Roberto Andrade,
  • Carlos Ayala,
  • Paulina Morillo

摘要

This chapter reframes AI security as a systemic and socio-technical challenge rather than a purely technical task. Moving beyond traditional notions of access control and component-level protection, it introduces resilience, context-aware governance, and multi-layered defense as foundational principles for securing AI systems. The concept of systemic assurance is presented as a framework for identifying emergent vulnerabilities arising from interdependencies, distributional drift, adversarial adaptation, and feedback loops in real-world deployments.