Confucius, Cyberpunk and Mr. Science: Comparing AI ethics between China and the EU

Different cultures prioritize different ethical questions about AI: Rather than simply disagreeing on answers, China and the EU tend to ask fundamentally different questions about AI ethics based on their distinct philosophical traditions and political systems, suggesting that a one-size-fits-all global approach to AI governance may be inadequate.

Shared principles can mask different underlying meanings: Even when China's and the EU's AI ethics frameworks appear to agree on common principles, they may interpret these concepts very differently based on their cultural heritage and values, meaning that surface-level consensus on AI ethics principles does not guarantee genuine alignment.

Current AI ethics codes show problematic homogeneity: The rapid proliferation of AI ethics codes from various stakeholders (industry, governments, international organizations) has created a landscape of presumed consensus that obscures important cultural and philosophical differences in how societies approach ethical issues.

Ethical pluralism requires understanding context, not just principles: The paper argues that meaningful AI ethics governance requires examining the philosophical roots and cultural contexts underlying different ethical frameworks, rather than simply adopting universal principles that may not reflect diverse moral doctrines and worldviews.

Western philosophical traditions are not universally applicable: Ethical systems like Kantian deontology, which aspire to universalism, are "not universal in practice," suggesting that AI ethics frameworks heavily influenced by Western philosophy may not adequately address the concerns and values of non-Western societies.