Recent advances in AI-based video synthesis and character generation technologies have opened new possibilities for content production. However, as AI-generated humanlike characters become increasingly realistic, they often evoke discomfort and rejection—a phenomenon known as the "uncanny valley." While technical realism has progressed rapidly, it alone does not guarantee emotional engagement or viewer acceptance. This study explores how emotional design, particularly as conceptualized by Donald Norman in visceral, behavioral, and reflective dimensions, serves as a strategic response to the uncanny valley in AI-generated content. Through case analyses—such as anthropomorphized animal shorts, interactive AI chatbots, and the reanimation of historical figures—this paper demonstrates how emotional design fosters trust, immersion, and affective resonance. Furthermore, it discusses the ethical implications of emotional manipulation and authenticity in AI-human interactions. By examining how emotional design restructures user perception beyond visual fidelity, this study offers a reinterpretation of the uncanny valley from an affective and aesthetic perspective.

ABSTRACT
1. 서론
2. 이론적 배경
    2.1 불쾌한 골짜기(Uncanny Valley)
    2.2 감정 설계(Emotional Design)
3. AI 콘텐츠에 적용된 감정 설계의 실제 사례
    3.1 본능적 감정 설계에 따른 AI 콘텐츠 사례
    3.2 행동적 감정 설계에 따른 AI 콘텐츠 사례
    3.3 반성적 감정 설계에 따른 AI 콘텐츠 사례
4. 감정 설계로 본 AI 콘텐츠의 수용 전략과 윤리
    4.1 감정 설계를 통한 불쾌한 골짜기의 재구성
    4.2 감정 설계의 윤리와 기술적 책임성
5. 결론
참고문헌

• Kim Soyoung(Kyungil University, 33 Buho-ri, Hayang-eup, Gyeongsan-si, Gyeongbuk, Republic of Korea) | 김소영 Corresponding author