認知迷失與現實脫節
Operation & Monitoring
Risk Description
Highly realistic AI-generated content (text, images, video, audio) blurs the boundary between real and fake, making it difficult for users to distinguish AI-generated content from genuine information. Long-term exposure to mixed true-false information environments may cause "cognitive disorientation" — users doubting the authenticity of all information, or conversely uncritically accepting all content. This phenomenon may undermine individual cognitive autonomy and societys consensus foundation.
Framework Mappings
iso 23894R6
tw principle透明與可解釋
tw principle人類自主
tw risk type社會系統性影響（Societal Systemic）
Controls English translation pending
- AI 內容標記標準
- 媒體素養教育
Implementation Steps English translation pending
- 推動 AI 生成內容的強制標記標準（如 C2PA）
- 開發並推廣 AI 內容真偽辨識工具
- 在教育系統中加強媒體素養和 AI 素養教育
- 支持建立可信賴的資訊來源驗證基礎設施