Anthropic Co-founder Urges External AI Oversight

Anthropic Co-founder Urges External AI Oversight

VATICAN CITY, May 25 () – Christopher Olah, co-founder of AI company Anthropic, stated on Monday that artificial intelligence development requires guidance beyond technology companies, advocating for increased oversight from religious leaders, governments, and civil society. Speaking at the Vatican during the presentation of Pope Leo XIV's first encyclical on AI, Olah emphasized the potential for large-scale displacement of human labor by AI.

Olah highlighted that AI companies operate under commercial, geopolitical, and personal pressures that may conflict with broader societal interests. He noted that even well-intentioned researchers are influenced by these forces, making external scrutiny essential. Anthropic, a U.S.-based developer of Claude AI tools, has previously implemented guardrails restricting military use of its models, such as autonomous targeting or domestic surveillance.

Olah welcomed the Catholic Church's engagement with AI, asserting that the ethical questions extend beyond engineering. He called for "earnest, thoughtful critics" to challenge companies and guide the development of powerful new systems positively. Olah identified three critical areas: the risk of widespread job losses, ensuring global distribution of AI benefits, and interpreting complex, opaque system behaviors.