[4Q3-OS-23] OS-23

[4Q3-OS-23-04] The Recreation of the Deceased by AI and Its Legal and Ethical Challenges

〇Kyoko Yoshinaga^1,2,3,4 (1. Project Associate Professor of the Graduate School of Media and Governance, Keio University, 2. Non-Resident Fellow, Institute for Technology Law & Policy, Georgetown Law Center, 3. Visiting Researcher, Institute for Future Initiatives, The University of Tokyo, 4. OECD-GPAI Expert)

[[Online]]

Keywords:generative AI, the privacy of the deceased, ethics

The rise of generative AI has enabled the "recreation" or "revival" of the deceased, but is this acceptable? In Japan, examples like "AI Hibari Misora" have emerged, while abroad, AI-generated appearances of Michael Jackson and projects recreating prominent judges have gained attention. In the U.S., AI services now allow users to converse with deceased family members based on their data from when they were alive.

Legally, Japanese copyright law protects rights for 70 years after death, and criminal law safeguards the honor of the deceased under certain conditions. However, the Act on the Protection of Personal Information generally excludes deceased individuals, and privacy rights are considered to end upon death.

Ethically, the inability of the deceased to defend against AI-generated images poses risks of misrepresentation. Android recreations may give the illusion of continued life. If such androids assume the deceased's societal roles, what implications arise for families, organizations, and society?

In this panel, I will discuss these legal and ethical challenges, drawing on global examples.