Abstract
The proliferation of applications of artificial intelligence(Al)raises fundamental questions about the extent to which human beings will tolerate Al-induced alterations to their subjective agency.While AI development should never compromise human dignity,contemporary implementations threaten three critical dimensions of this fundamental value:species-level supremacy,societal citizenship,and individual autonomy.This paper proposes a dual-prohibition regulatory paradigm to address these challenges.Absolute prohibition should apply to AI systems that instrumently exploit individuals without generating compensatory benefits for personal,social,or human welfare.Conditional restrictions should govern technologies presenting dignity risks alongside potential benefits.Under this risk-stratified framework,AI applications must undergo categorical classification based on their dignity infringement potentials.A proportionality assessment should then be applied to delineate operational boundaries through rigorous benefit-risk balancing.Specifically,legal mechanisms should:(i)impose stringent regulations on high-risk Al systems with substantial dignity implications;(i)implement measured governance for technologies that pose contained risks.Additionally,this paper advocates for legal reforms to codify fundamental human dignity protection into applicator's binding obligations.Regulatory compliance should be ensured through synergistic implementation of static proportionality evaluations and dynamic risk control mechanisms.
| Translated title of the contribution | The Boundaries and Regulation of Applications of Artificial Intelligence Based on Human Dignity |
|---|---|
| Original language | Chinese |
| Pages (from-to) | 32-46 |
| Number of pages | 0 |
| Journal | Journal of Comparative Law |
| Issue number | 2 |
| State | Published - 1 Apr 2025 |
Indexed by
- CSSCI
Fingerprint
Dive into the research topics of 'The Boundaries and Regulation of Applications of Artificial Intelligence Based on Human Dignity'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver