Anthropic co-founder discusses AI ethics after Pope Leo XIV’s encyclical

Anthropic co-founder Chris Olah warned that frontier AI development is increasingly shaped by commercial and geopolitical pressures. He said that during remarks delivered at the Vatican presentation of Pope Leo XIV’s new AI-focused encyclical Magnifica humanitas.

Speaking in Vatican City, Olah said advanced AI systems raise questions extending beyond computer science and engineering into ethics, philosophy, governance, and public policy. He argued that decisions surrounding AI systems and their societal impact should involve broader participation from public institutions and civil society rather than remaining concentrated within technology companies alone.

Olah also highlighted concerns about the social and economic effects of AI deployment, including possible labour-market disruption and unequal distribution of AI-related economic benefits. According to Olah, advanced AI development remains concentrated in a limited number of countries and organisations, while mechanisms for broader distribution of benefits remain unclear.

The remarks also addressed ongoing scientific uncertainty surrounding the internal behaviour of advanced AI systems. Olah said researchers continue to identify complex and not yet fully understood patterns within large AI models. He noted that some researchers have drawn comparisons between certain AI model behaviours and aspects of human cognition, while stressing the need for continued research and public scrutiny.

The remarks echoed themes from Pope Leo XIV’s encyclical Magnifica humanitas, which called for stronger safeguards around AI governance, accountability, and protection of human dignity.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!